Starting in version 2.0-0, OneMap can also deal with inbred-based populations, that is, populations that have homozygous parental lines in the genealogy (F2s, backcrosses and RILs). As a consequence, linkage phases do not need to be estimated.
In this vignette we explain how to proceed the analysis in an F2 population. The same procedure can be used for backcrosses and RILs as well, and therefore users should not have any difficulty in analyzing their data. However, there are a number of differences from genetic mapping in outcrossing species; please read the proper vignette.
If you are not familiar with R, we recommend first the reading of vignette Introduction to R. You do not need to be an expert in R to build your linkage map, but some concepts are necessary and will help you through the process.
There is a github OneMap version which is constantly improved, we strong recommend all users to try this version. In augusto-garcia/onemap github page you can find instructions to install the package from github and also more fancy tutorials.
For F2s, backcrosses and RILs, two input formats are accepted. The user can choose between the standard {} file format or the same raw file used by MAPMAKER/EXP (Lander et al., 1987). Therefore, one should have no difficulty in using data sets already available for MAPMAKER/EXP when deciding to try OneMap.
Both types of raw file can contain phenotypic information, but this will not be used during map construction, that requires only genotypic information (made available by molecular markers).
MAPMAKER/EXP data fileThe MAPMAKER/EXP raw file, combined with the map file produced by OneMap, can be readily used for QTL mapping using R/qtl (Broman et al., 2008) or QTL Cartographer (Wang et al., 2010), among others.
Here, we briefly present how to set up this data file. For more detailed information see the MAPMAKER/EXP manual (Lincon et al., 1993), available here.
The first line of your data file should be:
data type xxxx
where xxxx is one of the following data types:
xxxx |
Population type |
|---|---|
f2 backcross |
Backcross |
f2 intercross |
F2 |
ri self |
RIL, produced by selfing |
ri sib |
RIL, produced by sib mating |
The second line should contain the number of individuals in the progeny, the number of markers and the number of quantitative traits. So, for example, a valid line would be
10 5 2
for a data set with 10 individuals (yes, very small, but this is just an example), 5 markers and 2 traits evaluated.
Then, the genotype information is included for each marker. The character * indicates the beginning of information of a marker, followed by the marker name. For instance, here is an example of such a file for an F2 population with 10 individuals, 5 markers and 2 quantitative traits:
data type f2 intercross
10 5 2
*M1 A B H H A - B A A B
*M2 C - C C C - - C C A
*M3 D B D D - - B D D B
*M4 C C C - A C C A A C
*M5 C C C C C C C C C C
*weight 10.2 - 9.4 11.3 11.9 8.9 - 11.2 7.8 8.1
*length 1.7 2.1 - 1.8 2.0 1.0 - 1.7 1.0 1.1
The codification for genotypes is the following:
| Code | Meaning |
|---|---|
A |
homozygous for allele A (from parent 1 - AA) |
B |
homozygous for allele B (from parent 2 - BB) |
H |
heterozygous carrying both alleles (AB) |
C |
Not homozygous for allele A (Not AA) |
D |
Not homozygous for allele B (Not BB) |
- |
Missing data for the individual at this marker |
The symbols option (not included in this example), used in MAPMAKER/EXP files, is also accepted (please, see its manual for details).
The quantitative trait data should come after the genotypic data and has a similar format, except the trait values for each individual must be separated by at least one space, a tab or a line break. A dash (-) indicates missing data.
This file must be saved in plain text format using a simple text editor such as notepad on Microsoft Windows. Historically, MAPMAKER/EXP uses the .raw extension for this file; however, you can use any other extensions, such as .txt.
If you want to see more examples about this file type, open mapmaker_example_bc.raw and mapmaker_example_f2.raw, both available with OneMap and saved in the directory extdata on your computer, in the folder where you installed OneMap (use system.file(package="onemap") to see where it is located on your computer).
Now, let us load OneMap:
library(onemap)To save your project anytime, type:
save.image("C:/.../yourfile.RData")if you are using Windows; otherwise, adapt the code. Notice that you need to specify where to save and the name of the file. You can also use the toolbar, of course.
OneMap data fileThe OneMap data file has few differences compared to MAPMAKER/EXP format. As MAPMAKER/EXP format, the input OneMap file is a text file, where the first line indicates the cross type and the second line provides information about the number of individuals and the number of markers, but, because the format aldo supports to keep physical markers locations informations, the followed numbers indicate the presence/absence (1/0) of chromossome and position informations and, after the presence/absence(1/0) of phenotypic data.
The third line contains sample IDs. Then, the genotype information is included separately for each marker. The character * indicates the beginning of information input for a new marker, followed by the marker name. Next, there is a code indicating the marker type according with:
| Code | Type |
|---|---|
A.H.B |
Codominant marker |
C.A |
Dominant marker for allele B |
D.B |
Dominant marker for allele A |
Finally, after each marker type, comes the genotype data for the segregating population. Missing data are indicated with the character - (minus sign) and an empty space separates the information for each individual. Positions and phenotype information, if present, follows genotypic data with a similar structure. Details are found in the help of function read_onemap.
Here is an example of such file for 10 individuals and 5 markers (the three zeros in the second line indicate that there is no chromosome information, physical position information or phenotypic data, respectively). It is very similar to a MAPMAKER/EXP file, but has additional information about the crosstype.
data type f2 intercross
10 5 1 0 1
I1 I2 I3 I4 I5 I6 I7 I8 I9 I10
*M1 A.H.B ab a - ab b a ab - ab b
*M2 A.H.B a - ab ab - b a - a ab
*M3 C.A c a a c c - a c a c
*M4 A.H.B ab b - ab a b ab b - a
*M5 D.B b b d - b d b b b d
*CHROM 1 1 1 2 2 2 3 3 3 3
*fen1 10.3 11.2 11.1 - 9.8 8.9 11.0 10.7 - 10.1
*fen2 42 49 - 45 51 42 28 32 38 40
The input file must be saved in text format, with extensions like .raw. It is a good idea to open the text file called onemap_example_f2.raw (available in extdata with OneMap and saved in the directory you installed it) to see how this file should be. You can see where OneMap is installed using the command
OneMap data file from VCF fileIf you are working with biallelic markers, as SNPs and indels (only codominant markers A.H.B), in VCF (Variant Call Format) files, you can use the function vcf2raw to convert the genotype information from the VCF file to OneMap format .raw. Moreover, the function keeps chromosome and position information for each marker at the end of raw file.
To see how the function works, there is a short example of VCF file called vcf_example_f2.vcf saved in the directory you installed the OneMap. This file is already sorted, compressed and tabix indexed as require by the function. If you need to sort your file, we suggest the vcf-sort perl script from vcftools program. The functions bgzip and indexTabix from Rsamtools package can be used to compress and add tabix index.
vcf2raw(input = system.file("extdata/vcf_example_f2.vcf.gz", package = "onemap"),
output = "vcf_example_f21.raw", parent1 = "P1", parent2 = "P2", cross = "f2 intercross")The first argument is the path and the name of the data file. You can acess the example data using the function system.file as already seen above. For the second argument you can choose a name for the output .raw data. At the parent1 and parent2 arguments you should specify the parents samples IDs. They could be only one ID or, if there are replicate samples of the parents, multiple IDs for each parent. If you have replicates, you can specify a real number between 0.0 and 1.0 for the argument min_class to define the proportion of parents replicates that must have the same genotype for it to be assigned to the corresponding parent. At the cross argument you should point the cross type. You can access more information in the help page ?vcf2raw.
After running the function, the OneMap imput file vcf_example_out.raw generated will be at your working directory.
The function will not convert non-informative markers (if parents genotypes are not as expected). Therefore, you can have smaller number of markers in your .raw file compared to vcf file. In our example, there is one non-informative marker, identified as SNP4.
MAPMAKER/EXP fileOnce you created your data file with raw data, you can use OneMap function read_mapmaker to import it to OneMap:
mapmaker_example_f2 <- read_mapmaker(dir="C:/workingdirectory",
file="your_data_file.raw")The first argument is the directory where the input file is located, so modify it accordingly. The second one is the data file name.
In this example, an object named mapmaker_example_f2.raw was created. Notice that if you leave the argument dir blank, the file will be read from your current working directory. To set a working directory, see Introduction to R (Importing and Exporting Data).
mapmaker_example_f2 <- read_mapmaker(file= system.file("extdata/mapmaker_example_f2.raw",
package = "onemap"))For this example, we will use a simulated data set from an F2 population which is distributed along with OneMap. Because this particular data set is distributed along with the package, you can load it typing
data("mapmaker_example_f2")To see what this data set is about, type
mapmaker_example_f2
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 66
#> CHROM information: no
#> POS information: no
#> Percent genotyped: 85
#>
#> Segregation types:
#> AA : AB : BB --> 36
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> Trait_1: 0As you can see, the data consists of a sample of 200 individuals genotyped for 66 markers (36 co-dominant (AA, AB or BB), 15 dominant in one parent (Not AA or AA) and 15 dominant in the other parent (Not BB or BB) with 15% of missing data. You can also see that there is phenotypic information for one trait in the data set, that can be used for QTL mapping.
OneMap raw fileThe same procedure is made for OneMap raw file, but, instead of using the function read_mapmaker we use read_onemap to read the OneMap format.
onemap_example_f2 <- read_onemap(dir="C:/workingdirectory",
inputfile = "your_data_file.raw")In this example, an object named onemap_example_f2.raw was created. The data set containg the same markers and individual of the mapmaker_example_f2.raw file. Would be a good idea to open this two files in a text editor and compare them to better understand the differences between the two kinds of input files. We can read the onemap_example_f2.raw using:
onemap_example_f2 <- read_onemap(inputfile= system.file("extdata/onemap_example_f2.raw",
package = "onemap"))Or, because this particular data are available together with OneMap package:
data("onemap_example_f2")To see what this data set is about, type
onemap_example_f2
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 66
#> CHROM information: no
#> POS information: no
#> Percent genotyped: 85
#>
#> Segregation types:
#> AA : AB : BB --> 36
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> NA: 200As you can see, the mean difference in the output object is that read_onemap function keeps chromossome and position informations. Because the objetcs mapmaker_example_f2 and onemap_example_f2 are pratically the same, from now we will use only onemap_example_f2.
OneMap data sets from VCF file can also be read by read_onemap function, once the file have the OneMap format:
vcf_example_f2 <- read_onemap(inputfile= system.file("extdata/vcf_example_f2.raw",
package = "onemap"))This data is also available together with OneMap package:
data("vcf_example_f2")This file contains 25 other markers for 192 of the individuals in onemap_example_f2:
vcf_example_f2
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 192
#> No. markers: 25
#> CHROM information: yes
#> POS information: yes
#> Percent genotyped: 86
#>
#> Segregation types:
#> AA : AB : BB --> 25
#>
#> No. traits: 0Before building your linkage map, you should take a look at your data set. First, notice that by reading the raw data into OneMap, an object of classes onemap and f2 was produced:
class(onemap_example_f2)
#> [1] "onemap" "f2"
class(vcf_example_f2)
#> [1] "onemap" "f2"In fact, functions read_mapmaker and read_onemap will produce objects of classes backcross, riself, risib or f2, according to the information in the data file for inbred-based populations. Therefore, you can use OneMap’s version of function plot to produce a graphic with information about the raw data. It will automatically recognize the class of the object and produce the graphic. To see it in action, try:
plot(onemap_example_f2)plot(vcf_example_f2)The graphic is self-explanatory. If you want to save it, see the help for function plot.onemap:
?plot.onemapThis graphic shows that missing data is somehow randomly distributed; also, the proportion of dominant markers is relatively high for this data set. In Onemap’s notation, co-dominant markers are classified as of B type; dominant ones, by C type (for details about this notation, see the vignette for outcrossing species). You can see the number of loci within each type using funcion plot_by_segreg_type:
plot_by_segreg_type(onemap_example_f2)plot_by_segreg_type(vcf_example_f2)So, as shown before, the object onemap_example_f2 has 36 codominant markers and 30 dominant ones and the vcf_example_f2has only codominant markers.
OneMap objectsIf you have more than one dataset of markers, all from the same cross type, you can use the function combine_onemap to merge them into only one onemap object.
In our example, we have two onemap objects:
onemap_example_f2 (equivalent to mapmaker_example_f2) with 66 markers and 200 individualsvcf_example_f2 with 25 biallelic markers and 192 individuals.The combine_function recognizes the correspondent individuals by the ID, thus, it is important define exactly same IDs to respective individuals in both raw files. Compared with the first file, the second file do not have markers informations for 8 individuals. The combine_onemap will complete those informations with NA.
In our examples, we have only genotypic information, but the function can also merge the phenotipic information if it exists.
comb_example <- combine_onemap(onemap_example_f2, vcf_example_f2)
comb_example
#> This is an object of class 'onemap'
#> Type of cross: f2
#> No. individuals: 200
#> No. markers: 91
#> CHROM information: yes
#> POS information: yes
#> Percent genotyped: 84
#>
#> Segregation types:
#> AA : AB : BB --> 61
#> Not AA : AA --> 15
#> Not BB : BB --> 15
#>
#> No. traits: 1
#> Missing trait values:
#> NA: 200The function arguments are the names of the onemap objects you want to combine.
Plotting markers genotypes from the outputted onemap object, we can see that there are more missing data - (black vertical lines) for some individuals, because they were missing in the second file.
plot(comb_example)Now, it should be interesting to see if markers are segregating following what is expected by Mendel’s law. You first need to use function test_segregation using as argument an object of class onemap.
f2_test <- test_segregation(comb_example)This will produce an object of class onemap_segreg_test:
class(f2_test)
#> [1] "onemap_segreg_test"You cannot see the results if you simply type the object name; use Onemap’s version of the print function for objects of class onemap_segreg_test:
f2_test(Nothing is shown!)
print(f2_test)
#> Marker H0 Chi-square p-value % genot.
#> 1 M1 1:2:1 0.206896552 0.901722662 87.0
#> 2 M2 3:1 0.292682927 0.588506355 82.0
#> 3 M3 3:1 0.124031008 0.724703003 86.0
#> 4 M4 3:1 0.292682927 0.588506355 82.0
#> 5 M5 3:1 0.159763314 0.689374522 84.5
#> 6 M6 3:1 0.620689655 0.430791121 87.0
#> 7 M7 1:2:1 3.185185185 0.203397600 81.0
#> 8 M8 3:1 0.165644172 0.684012345 81.5
#> 9 M9 3:1 0.126984127 0.721579725 84.0
#> 10 M10 1:2:1 1.721893491 0.422761645 84.5
#> 11 M11 3:1 1.390946502 0.238245323 81.0
#> 12 M12 1:2:1 1.678160920 0.432107681 87.0
#> 13 M13 3:1 0.154285714 0.694472964 87.5
#> 14 M14 1:2:1 5.337209302 0.069348924 86.0
#> 15 M15 3:1 0.403292181 0.525393917 81.0
#> 16 M16 1:2:1 2.036144578 0.361290733 83.0
#> 17 M17 1:2:1 1.000000000 0.606530660 81.0
#> 18 M18 1:2:1 3.413793103 0.181427972 87.0
#> 19 M19 3:1 0.279069767 0.597311573 86.0
#> 20 M20 3:1 4.462626263 0.034644194 82.5
#> 21 M21 1:2:1 4.892857143 0.086602329 84.0
#> 22 M22 1:2:1 2.862068966 0.239061489 87.0
#> 23 M23 3:1 0.008032129 0.928587488 83.0
#> 24 M24 3:1 8.649325626 0.003271824 86.5
#> 25 M25 1:2:1 0.686746988 0.709373215 83.0
#> 26 M26 1:2:1 7.022598870 0.029858091 88.5
#> 27 M27 1:2:1 0.108433735 0.947226662 83.0
#> 28 M28 1:2:1 3.319526627 0.190183989 84.5
#> 29 M29 1:2:1 0.161676647 0.922342801 83.5
#> 30 M30 1:2:1 0.607142857 0.738177160 84.0
#> 31 M31 3:1 0.007843137 0.929430415 85.0
#> 32 M32 3:1 0.016949153 0.896416968 88.5
#> 33 M33 1:2:1 0.640718563 0.725888192 83.5
#> 34 M34 3:1 0.235867446 0.627206926 85.5
#> 35 M35 3:1 0.606741573 0.436017310 89.0
#> 36 M36 3:1 0.272727273 0.601508134 88.0
#> 37 M37 1:2:1 3.169491525 0.204999905 88.5
#> 38 M38 1:2:1 3.055555556 0.217017393 90.0
#> 39 M39 3:1 0.163636364 0.685830434 82.5
#> 40 M40 1:2:1 4.872093023 0.087506123 86.0
#> 41 M41 1:2:1 4.253012048 0.119253235 83.0
#> 42 M42 3:1 0.124031008 0.724703003 86.0
#> 43 M43 1:2:1 0.927272727 0.628992237 82.5
#> 44 M44 1:2:1 2.000000000 0.367879441 86.0
#> 45 M45 3:1 0.008032129 0.928587488 83.0
#> 46 M46 1:2:1 0.740112994 0.690695307 88.5
#> 47 M47 1:2:1 0.758620690 0.684333200 87.0
#> 48 M48 1:2:1 2.122807018 0.345969898 85.5
#> 49 M49 3:1 1.190476190 0.275233524 87.5
#> 50 M50 3:1 1.068686869 0.301242240 82.5
#> 51 M51 3:1 0.031746032 0.858586201 84.0
#> 52 M52 1:2:1 2.684848485 0.261211660 82.5
#> 53 M53 3:1 0.017142857 0.895830102 87.5
#> 54 M54 1:2:1 0.772455090 0.679615865 83.5
#> 55 M55 1:2:1 0.655172414 0.720661163 87.0
#> 56 M56 1:2:1 2.310734463 0.314941859 88.5
#> 57 M57 1:2:1 6.159763314 0.045964696 84.5
#> 58 M58 3:1 2.050847458 0.152121494 88.5
#> 59 M59 3:1 0.074074074 0.785494747 81.0
#> 60 M60 3:1 0.050505051 0.822186767 82.5
#> 61 M61 1:2:1 0.655172414 0.720661163 87.0
#> 62 M62 1:2:1 1.047337278 0.592343463 84.5
#> 63 M63 1:2:1 2.147928994 0.341651353 84.5
#> 64 M64 3:1 0.238658777 0.625176499 84.5
#> 65 M65 1:2:1 4.571428571 0.101701392 84.0
#> 66 M66 1:2:1 0.452513966 0.797513128 89.5
#> 67 SNP1 1:2:1 1.278787879 0.527612092 82.5
#> 68 SNP2 1:2:1 1.559523810 0.458515169 84.0
#> 69 SNP3 1:2:1 0.503105590 0.777592404 80.5
#> 70 SNP4 1:2:1 0.165644172 0.920514905 81.5
#> 71 SNP5 1:2:1 4.450000000 0.108067419 80.0
#> 72 SNP6 1:2:1 2.641975309 0.266871595 81.0
#> 73 SNP7 1:2:1 3.745341615 0.153712576 80.5
#> 74 SNP8 1:2:1 4.164705882 0.124636604 85.0
#> 75 SNP9 1:2:1 2.578313253 0.275503037 83.0
#> 76 SNP10 1:2:1 3.359281437 0.186440949 83.5
#> 77 SNP11 1:2:1 2.395061728 0.301938820 81.0
#> 78 SNP12 1:2:1 4.731707317 0.093869134 82.0
#> 79 SNP13 1:2:1 2.963855422 0.227199291 83.0
#> 80 SNP14 1:2:1 7.123456790 0.028389714 81.0
#> 81 SNP15 1:2:1 3.109090909 0.211285400 82.5
#> 82 SNP16 1:2:1 2.751552795 0.252643368 80.5
#> 83 SNP17 1:2:1 5.850299401 0.053656659 83.5
#> 84 SNP18 1:2:1 3.108433735 0.211354837 83.0
#> 85 SNP19 1:2:1 3.335403727 0.188680181 80.5
#> 86 SNP20 1:2:1 3.867469880 0.144607090 83.0
#> 87 SNP21 1:2:1 3.238095238 0.198087264 84.0
#> 88 SNP22 1:2:1 4.455621302 0.107764105 84.5
#> 89 SNP23 1:2:1 2.847058824 0.240862412 85.0
#> 90 SNP24 1:2:1 1.778443114 0.410975549 83.5
#> 91 SNP25 1:2:1 1.981595092 0.371280460 81.5This shows the results of the Chi-square test for the expected Mendelian segregation pattern of each marker locus. This depends of course on marker type, because co-dominant markers can show heterozygous genotypes. The appropriate null hypothesis is selected by the function. The proportion of individuals genotyped is also shown.
To declare statistical significance, remember that you should consider that multiple tests are being performed. To guide you in the analysis, function Bonferroni_alpha shows the alpha value that should be considered for this number of loci if applying Bonferroni’s correction with global alpha of 0.05:
Bonferroni_alpha(f2_test)
#> [1] 0.0005494505You can subset object f2_test to see which markers are distorted under Bonferroni’s criterion, but it is easier to see the proportion of markers that are distorted by drawing a graphic using OneMap’s version of the function plot for objects of class onemap_segreg_test:
plot(f2_test)The graphic is self-explanatory: p-values were transformed by using -log10(p-values) for better visualization. A vertical line shows the threshold for tests if Bonferroni’s correction is applied. Significant and non-significant tests are identified. In this particular example, no test was statistically significant, so none will be discarded.
Please, remember that Bonferroni’s correction is conservative, and also that discarding marker data might not be a good approach to your analysis. This graphic is just to suggest a criterion, so use it with caution.
You can see a list of markers with non-distorted segregation using function select_segreg:
select_segreg(f2_test)
#> [1] "M1" "M2" "M3" "M4" "M5" "M6" "M7" "M8"
#> [9] "M9" "M10" "M11" "M12" "M13" "M14" "M15" "M16"
#> [17] "M17" "M18" "M19" "M20" "M21" "M22" "M23" "M24"
#> [25] "M25" "M26" "M27" "M28" "M29" "M30" "M31" "M32"
#> [33] "M33" "M34" "M35" "M36" "M37" "M38" "M39" "M40"
#> [41] "M41" "M42" "M43" "M44" "M45" "M46" "M47" "M48"
#> [49] "M49" "M50" "M51" "M52" "M53" "M54" "M55" "M56"
#> [57] "M57" "M58" "M59" "M60" "M61" "M62" "M63" "M64"
#> [65] "M65" "M66" "SNP1" "SNP2" "SNP3" "SNP4" "SNP5" "SNP6"
#> [73] "SNP7" "SNP8" "SNP9" "SNP10" "SNP11" "SNP12" "SNP13" "SNP14"
#> [81] "SNP15" "SNP16" "SNP17" "SNP18" "SNP19" "SNP20" "SNP21" "SNP22"
#> [89] "SNP23" "SNP24" "SNP25"To get a list of distorted ones (none in this example):
select_segreg(f2_test, distorted = TRUE)
#> character(0)For the next steps will be useful to know the numbers of each markers with segregation distortion, so then you can keep those out of your map building analysis. These numbers refer to the lines where markers are located on the data file.
To access the corresponding number for of this markers you can change the numbers argument:
no_dist <- select_segreg(f2_test, distorted = FALSE, numbers = TRUE) #to show the markers numbers without segregation distortion
no_dist
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
#> [24] 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
#> [47] 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
#> [70] 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
dist <- select_segreg(f2_test, distorted = TRUE, numbers = TRUE) #to show the markers numbers with segregation distortion
dist
#> integer(0)After visualizing raw data and checking for segregation distortion, let us now estimate recombination fractions between all pairs of markers (two-point tests). This is necessary to allow us to test which markers are linked. At this point, you should pay no attention if markers show segregation distortion or not, that is, simply use all of them.
twopts_f2 <- rf_2pts(comb_example)There are two optional arguments in function rf_2pts: LOD and max.rf which indicate the minimum LOD Score and the maximum recombination fraction to declare linkage (they default to 3.0 and 0.5, respectively).
The default for the recombination fraction is easy to understand, because if max.rf < 0.5 we could state that markers are linked. The LOD Score is the statistic used to evaluate the significance of the test for max.rf = 0.50. This needs to take into consideration the number of tests performed, which of course depends on the number of markers. Function suggest_lod can help users to find an initial value to use for their linkage test. For this example:
(LOD_sug <- suggest_lod(comb_example))
#> [1] 4.155071Thus, one should consider using LOD = 4.155 for the tests. Please, notice that this is just a guide, not a value to take without any further consideration. For now, we will keep the default values, but later will show that results do not change in our example by using LOD = 3 or LOD = 4.155.
If you want to see the results for a single pair of markers, say M12 and M42, use:
print(twopts_f2, c("M12", "M42"))
#> Results of the 2-point analysis for markers: M12 and M42
#> Criteria: LOD = 3 , Maximum recombination fraction = 0.5
#>
#> rf LOD
#> 0.3575533 1.9430940This was possible because OneMap has a version of the print function that can be applied to objects of class rf_2pts:
class(twopts_f2)
#> [1] "rf_2pts" "f2"However, objects of this type are too complex to print if you do not specify a pair of markers:
print(twopts_f2)
#> This is an object of class 'rf_2pts'
#>
#> Criteria: LOD = 3 , Maximum recombination fraction = 0.5
#>
#> This object is too complex to print
#> Type 'print(object, c(mrk1=marker, mrk2=marker))' to see
#> the analysis for two markers
#> mrk1 and mrk2 can be the names or numbers of both markersIn this example we follow two different strategies:
Using only recombinations informations.
Using the recombinations and also the reference genome informations, once our example have CHROM and POS informations for some of the markers.
First, we will apply the strategy using only recombinations informations. In a second part of this tutorial we show a way to use also reference genome informations. At the and of our analysis we will be able to compare this two strategies drawing the resulted genetic maps.
To assign markers to linkage groups, first use the function make_seq to create a (un-ordered) sequence with all markers:
mark_all_f2 <- make_seq(twopts_f2, "all")Function make_seq is used to create sequences from objects of several different classes. Here, the first argument is of class rf_2pts and the second argument specifies which markers one wants to use ("all" indicates that all markers will be analyzed). The object mark_all_f2 is of class sequence:
class(mark_all_f2)
#> [1] "sequence"If you want to form groups with a subset of markers, say M1, M3 and M7, use:
mrk_subset <- make_seq(twopts_f2, c(1, 3, 7))In this case, it was easy because marker names and order in the objects (indicated in vector c(1, 3, 7)) are closed related, that is, you can easily know the position of markers in the object once you know their names. However, this is not true for real data sets, where markers do not have simple names such as M1 or M2.
A good example is to use the vector of markers without segregation distortion that we selected when applying the Chi-square tests.
mark_no_dist_f2 <- make_seq(twopts_f2, no_dist)In our example, there are no markers with segretion distortion, then the object mark_no_dist_f2 is equivalent to mark_all_f2.
You can assign markers to linkage groups using the function group:
LGs_f2 <- group(mark_all_f2)
#> Selecting markers:
#> group 1
#> .....................................
#> group 2
#> ..................
#> group 3
#> ....................................
LGs_f2
#> This is an object of class 'group'
#> It was generated from the object "mark_all_f2"
#>
#> Criteria used to assign markers to groups:
#> LOD = 3 , Maximum recombination fraction = 0.5
#>
#> No. markers: 91
#> No. groups: 3
#> No. linked markers: 91
#> No. unlinked markers: 0
#>
#> Printing groups:
#> Group 1 : 37 markers
#> M1 M3 M4 M6 M7 M9 M12 M13 M17 M23 M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58 M60 M63 SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10
#>
#> Group 2 : 18 markers
#> M2 M5 M8 M10 M11 M25 M32 M33 M37 M41 M43 M45 M51 M54 M61 M66 SNP24 SNP25
#>
#> Group 3 : 36 markers
#> M14 M15 M16 M18 M19 M20 M21 M22 M24 M28 M38 M39 M47 M48 M49 M50 M52 M56 M57 M59 M62 M64 M65 SNP11 SNP12 SNP13 SNP14 SNP15 SNP16 SNP17 SNP18 SNP19 SNP20 SNP21 SNP22 SNP23This will show the linkage groups that are formed if criteria defined by max.rf and LOD are applied to define thresholds to be used when assigning markers to linkage groups. If not modified, the same values used for the object twopts (from two-point analysis) will be maintained (so, LOD = 3.0 and max.rf = 0.5 in this example).
Users can easily change the default values. For example, using LOD suggested by suggest_lod (rounded up):
(LGs_f2 <- group(mark_all_f2, LOD = LOD_sug, max.rf = 0.5))
#> Selecting markers:
#> group 1
#> .....................................
#> group 2
#> ..................
#> group 3
#> ....................................
#> This is an object of class 'group'
#> It was generated from the object "mark_all_f2"
#>
#> Criteria used to assign markers to groups:
#> LOD = 4.155071 , Maximum recombination fraction = 0.5
#>
#> No. markers: 91
#> No. groups: 3
#> No. linked markers: 91
#> No. unlinked markers: 0
#>
#> Printing groups:
#> Group 1 : 37 markers
#> M1 M3 M4 M6 M7 M9 M12 M13 M17 M23 M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58 M60 M63 SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10
#>
#> Group 2 : 18 markers
#> M2 M5 M8 M10 M11 M25 M32 M33 M37 M41 M43 M45 M51 M54 M61 M66 SNP24 SNP25
#>
#> Group 3 : 36 markers
#> M14 M15 M16 M18 M19 M20 M21 M22 M24 M28 M38 M39 M47 M48 M49 M50 M52 M56 M57 M59 M62 M64 M65 SNP11 SNP12 SNP13 SNP14 SNP15 SNP16 SNP17 SNP18 SNP19 SNP20 SNP21 SNP22 SNP23No difference was obtained in this example. (The parentheses above are just to avoid typing LGs_f2 in a new row to have the object printed).
We can see that the markers were assigned to three linkage groups with 27, 16 and 23 markers, with no unlinked markers.
Notice the class of object LGs_f2:
class(LGs_f2)
#> [1] "group"After assignin markers to linkage groups, the next step is to order the markers within each group.
First, let us choose the mapping function used to display the genetic map. We can choose between Kosambi or Haldane mapping functions. To use Haldane, type
set_map_fun(type = "haldane")To use Kosambi’s function:
set_map_fun(type = "kosambi")We will start our analysis using the smallest linkage group (group 2). To do so, you must extract it from the object of class group. Let us extract the group 2 using function make_seq:
LG2_f2 <- make_seq(LGs_f2, 2)The first argument is an object of class group and the second is a number indicating which linkage group will be extracted. In this case, the object LGs_f2, generated by function group, is of class group. In fact, this function can handle different classes of objects.
If you type
LG2_f2
#>
#> Number of markers: 18
#> Markers in the sequence:
#> M2 M5 M8 M10 M11 M25 M32 M33 M37 M41 M43 M45 M51 M54 M61 M66 SNP24 SNP25
#>
#> Parameters not estimated.you will see which markers are comprised in the sequence. But notice that no parameters have been estimated so far (the function says Parameters not estimated). This refers to the fact that so far we only attributed markers to linkage groups, but we did not perform any analysis for them as a group - only as pairs. (Does it seem complicated? Do not worry, you will understand details in a moment).
Notice the class of object LG2_f2:
class(LG2_f2)
#> [1] "sequence"To order markers in this group, you can use a two-point based algorithm such as Seriation (Buetow and Chakravarti, 1987), Rapid Chain Delineation (Doerge, 1996), Recombination Counting and Ordering (Van Os et al., 2005) and Unidirectional Growth (Tan and Fu, 2006):
LG2_ser_f2 <- seriation(LG2_f2)
LG2_rcd_f2 <- rcd(LG2_f2)
LG2_rec_f2 <- record(LG2_f2)
LG2_ug_f2 <- ug(LG2_f2)For this particular data set, the algorithms provided different results (results not printed in this vignette). For an evaluation and comparison of these methods, see Mollinari et al. (2009).
When possible (i.e., when groups have a small number of markers, in general up to 10 or 11), one should select the best order by comparing the multipoint likelihood of all possible orders between markers (exhaustive search). This procedure is implemented in the function compare. Although feasible for up to 10 or 11 markers, with 7 or more markers it will take a couple of hours until you see the results (depending of course on the computational resources available).
Linkage group 2 has 18 markers, so using function compare is infeasible. Thus we will apply a heuristic that shows reliable results. First, we will choose a moderate number of markers, say 6, to create a framework using the function compare, and then we will position the remaining markers into this framework using function try_seq. The way we choose these initial markers in inbred-based populations is somewhat different from what we did for outcrossing populations, where there is a mixture of segregation patterns (see the vignette for details).
In our scenario, we recommend two methods:
compare). If the LOD Score of the second best order is greater than a given threshold, say, 3, then take the best order to proceed with the next step. If not, repeat the procedure.compare. Next, try to map the remaining markers, one at a time, beginning with co-dominant ones (most informative ones), then add the dominant ones.You can do this procedure manually, in a similar way as done for outcrossing species (see the vignette for details). However, this procedure is automated in function order_seq, which we will use here:
LG2_f2_ord <- order_seq(input.seq = LG2_f2, n.init = 5,
subset.search = "twopt",
twopt.alg = "rcd", THRES = 3)
#>
#> Cross type: f2
#> Choosing initial subset using 'two-point' approach
#>
#> order obtained using RCD algorithm:
#>
#> 41 10 43 11 2 32 45 54 66 61 25 5 51 33 91 37 8 90
#>
#> calculating multipoint map using tol = 0.1 .
#>
#>
#> Comparing 60 orders:
#>
#>
|
| | 0%
|
|= | 2%
|
|== | 3%
|
|=== | 5%
|
|==== | 7%
|
|===== | 8%
|
|====== | 10%
|
|======== | 12%
|
|========= | 13%
|
|========== | 15%
|
|=========== | 17%
|
|============ | 18%
|
|============= | 20%
|
|============== | 22%
|
|=============== | 23%
|
|================ | 25%
|
|================= | 27%
|
|================== | 28%
|
|==================== | 30%
|
|===================== | 32%
|
|====================== | 33%
|
|======================= | 35%
|
|======================== | 37%
|
|========================= | 38%
|
|========================== | 40%
|
|=========================== | 42%
|
|============================ | 43%
|
|============================= | 45%
|
|============================== | 47%
|
|=============================== | 48%
|
|================================ | 50%
|
|================================== | 52%
|
|=================================== | 53%
|
|==================================== | 55%
|
|===================================== | 57%
|
|====================================== | 58%
|
|======================================= | 60%
|
|======================================== | 62%
|
|========================================= | 63%
|
|========================================== | 65%
|
|=========================================== | 67%
|
|============================================ | 68%
|
|============================================== | 70%
|
|=============================================== | 72%
|
|================================================ | 73%
|
|================================================= | 75%
|
|================================================== | 77%
|
|=================================================== | 78%
|
|==================================================== | 80%
|
|===================================================== | 82%
|
|====================================================== | 83%
|
|======================================================= | 85%
|
|======================================================== | 87%
|
|========================================================= | 88%
|
|========================================================== | 90%
|
|============================================================ | 92%
|
|============================================================= | 93%
|
|============================================================== | 95%
|
|=============================================================== | 97%
|
|================================================================ | 98%
|
|=================================================================| 100%
#>
#>
#>
#> Running try algorithm
#> 66 --> M66 : ......
#> 37 --> M37 : ......
#> 10 --> M10 : ......
#> 54 --> M54 : .......
#> 25 --> M25 : ........
#> 43 --> M43 : .........
#> 91 --> SNP25 : ..........
#> 32 --> M32 : ..........
#> 5 --> M5 : ..........
#> 51 --> M51 : ...........
#> 45 --> M45 : ............
#> 8 --> M8 : .............
#> 11 --> M11 : ..............
#>
#> LOD threshold = 3
#>
#> Positioned markers: 2 10 41 43 45 54 61 25 5 51 33 90 8
#>
#> Markers not placed on the map: 11 32 37 66 91
#>
#>
#> Calculating LOD-Scores
#> 11 --> M11 : ..............
#> 32 --> M32 : ..............
#> 37 --> M37 : ..............
#> 66 --> M66 : ..............
#> 91 --> SNP25 : ..............
#>
#>
#> Placing remaining marker(s) at most likely position
#> 37 --> M37 : ..............
#> 66 --> M66 : ...............
#> 11 --> M11 : ................
#> 91 --> SNP25 : .................
#> 32 --> M32 : ..................
#>
#> Estimating final genetic map using tol = 10E-5.The first argument is an object of class sequence (LG2_f2). n.init = 5 means that five markers will be used in the compare step. The argument subset.search = "twopt" indicates that these five markers should be chosen by using a two point method, which will be Rapid Chain Delineation, as indicated by the argument twopt.alg = "rcd". THRES = 3 indicates that the try_seq step will only add markers to the sequence which can be mapped with LOD Score greater than 3.
Check the order obtained by this procedure:
LG2_f2_ord
#>
#> Best sequence found.
#> Printing map:
#>
#> Markers Position
#>
#> 2 M2 0.00
#> 10 M10 7.06
#> 41 M41 14.96
#> 43 M43 31.26
#> 45 M45 38.96
#> 54 M54 41.02
#> 61 M61 49.73
#> 25 M25 55.66
#> 5 M5 59.81
#> 51 M51 64.09
#> 33 M33 70.24
#> 90 SNP24 75.07
#> 8 M8 92.65
#>
#> 13 markers log-likelihood: -1005.635
#>
#>
#>
#> The following markers could not be uniquely positioned.
#> Printing most likely positions for each unpositioned marker:
#>
#> ------------------------------------
#> | | 11 | 32 | 37 | 66 | 91 |
#> |----|-----|-----|-----|-----|-----|
#> | | | | | | |
#> | 2 | | | | | |
#> | | * | | | | |
#> | 10 | | | | | |
#> | | | | | | |
#> | 41 | | | | | |
#> | | *** | * | | | |
#> | 43 | | | | | |
#> | | | *** | | | |
#> | 45 | | | | | |
#> | | | ** | | | |
#> | 54 | | | | | |
#> | | | | | *** | |
#> | 61 | | | | | |
#> | | | | | | |
#> | 25 | | | | | |
#> | | | | | | |
#> | 5 | | | | | |
#> | | | | | | |
#> | 51 | | | | | |
#> | | | | | | *** |
#> | 33 | | | | | |
#> | | | | | | * |
#> | 90 | | | | | |
#> | | | | *** | | |
#> | 8 | | | | | |
#> | | | | | | |
#> ------------------------------------
#>
#> '***' indicates the most likely position(s) (LOD = 0.0)
#>
#> '**' indicates very likely positions (LOD > -1.0)
#>
#> '*' indicates likely positions (LOD > -2.0)Note that markers 11, 32, 37, 66 and 91 could not be safely mapped to a single position (LOD Score > THRES in absolute value). The output displays the safe order and the most likely positions for markers not mapped, where *** indicates the most likely position, and * corresponds to other plausible positions. (If you are familiar with MAPMAKER/EXP, you will recognize the representation).
To get the safe order, use
LG2_f2_safe <- make_seq(LG2_f2_ord, "safe")and to get the order with all markers (i.e., including the ones not mapped to a single position), use:
(LG2_f2_all <- make_seq(LG2_f2_ord, "force"))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 2 M2 0.00
#> 10 M10 7.06
#> 41 M41 14.75
#> 11 M11 27.41
#> 43 M43 32.04
#> 32 M32 35.93
#> 45 M45 41.52
#> 54 M54 43.89
#> 66 M66 50.91
#> 61 M61 52.58
#> 25 M25 58.29
#> 5 M5 62.26
#> 51 M51 66.71
#> 91 SNP25 74.56
#> 33 M33 75.93
#> 90 SNP24 80.81
#> 37 M37 89.22
#> 8 M8 95.02
#>
#> 18 markers log-likelihood: -1170.482which places markers 11, 32, 37, 66 and 91 into their most likely positions.
Although some old publications presented maps with only safe orders, we see no reason not to use option force, and recommend it for users.
The order_seq function can perform two rounds of the try_seq step, first using THRES and then THRES - 1 as the threshold. This generally results in safe orders with more markers mapped, but takes longer to run. To do this, type:
LG2_f2_ord <- order_seq(input.seq = LG2_f2, n.init = 5,
subset.search = "twopt",
twopt.alg = "rcd", THRES = 3,
touchdown = TRUE)
#>
#> Cross type: f2
#> Choosing initial subset using 'two-point' approach
#>
#> order obtained using RCD algorithm:
#>
#> 41 10 43 11 2 32 45 54 66 61 25 5 51 33 91 37 8 90
#>
#> calculating multipoint map using tol = 0.1 .
#>
#>
#> Comparing 60 orders:
#>
#>
|
| | 0%
|
|= | 2%
|
|== | 3%
|
|=== | 5%
|
|==== | 7%
|
|===== | 8%
|
|====== | 10%
|
|======== | 12%
|
|========= | 13%
|
|========== | 15%
|
|=========== | 17%
|
|============ | 18%
|
|============= | 20%
|
|============== | 22%
|
|=============== | 23%
|
|================ | 25%
|
|================= | 27%
|
|================== | 28%
|
|==================== | 30%
|
|===================== | 32%
|
|====================== | 33%
|
|======================= | 35%
|
|======================== | 37%
|
|========================= | 38%
|
|========================== | 40%
|
|=========================== | 42%
|
|============================ | 43%
|
|============================= | 45%
|
|============================== | 47%
|
|=============================== | 48%
|
|================================ | 50%
|
|================================== | 52%
|
|=================================== | 53%
|
|==================================== | 55%
|
|===================================== | 57%
|
|====================================== | 58%
|
|======================================= | 60%
|
|======================================== | 62%
|
|========================================= | 63%
|
|========================================== | 65%
|
|=========================================== | 67%
|
|============================================ | 68%
|
|============================================== | 70%
|
|=============================================== | 72%
|
|================================================ | 73%
|
|================================================= | 75%
|
|================================================== | 77%
|
|=================================================== | 78%
|
|==================================================== | 80%
|
|===================================================== | 82%
|
|====================================================== | 83%
|
|======================================================= | 85%
|
|======================================================== | 87%
|
|========================================================= | 88%
|
|========================================================== | 90%
|
|============================================================ | 92%
|
|============================================================= | 93%
|
|============================================================== | 95%
|
|=============================================================== | 97%
|
|================================================================ | 98%
|
|=================================================================| 100%
#>
#>
#>
#> Running try algorithm
#> 66 --> M66 : ......
#> 37 --> M37 : ......
#> 10 --> M10 : ......
#> 54 --> M54 : .......
#> 25 --> M25 : ........
#> 43 --> M43 : .........
#> 91 --> SNP25 : ..........
#> 32 --> M32 : ..........
#> 5 --> M5 : ..........
#> 51 --> M51 : ...........
#> 45 --> M45 : ............
#> 8 --> M8 : .............
#> 11 --> M11 : ..............
#>
#> LOD threshold = 3
#>
#> Positioned markers: 2 10 41 43 45 54 61 25 5 51 33 90 8
#>
#> Markers not placed on the map: 11 32 37 66 91
#>
#>
#>
#> Trying to map remaining markers with LOD threshold 2
#> 11 --> M11 : ..............
#> 32 --> M32 : ..............
#> 37 --> M37 : ..............
#> 66 --> M66 : ...............
#> 91 --> SNP25 : ................
#>
#> LOD threshold = 2
#>
#> Positioned markers: 2 10 41 43 45 54 66 61 25 5 51 33 90 37 8
#>
#> Markers not placed on the map: 11 32 91
#>
#>
#> Calculating LOD-Scores
#> 11 --> M11 : ................
#> 32 --> M32 : ................
#> 91 --> SNP25 : ................
#>
#>
#> Placing remaining marker(s) at most likely position
#> 11 --> M11 : ................
#> 91 --> SNP25 : .................
#> 32 --> M32 : ..................
#>
#> Estimating final genetic map using tol = 10E-5.The output is too big to be included here, so please try it to see what happens. In short, for this particular sequence, the touchdown step could additionally map two of the five markers (37 and 66), but this depends on the dataset. Let us continue our analysis using the order with all markers as suggested by the function order_seq:
(LG2_f2_final <- make_seq(LG2_f2_ord, "force"))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 2 M2 0.00
#> 10 M10 7.06
#> 41 M41 14.75
#> 11 M11 27.41
#> 43 M43 32.04
#> 32 M32 35.93
#> 45 M45 41.52
#> 54 M54 43.89
#> 66 M66 50.91
#> 61 M61 52.58
#> 25 M25 58.29
#> 5 M5 62.26
#> 51 M51 66.71
#> 91 SNP25 74.56
#> 33 M33 75.93
#> 90 SNP24 80.81
#> 37 M37 89.22
#> 8 M8 95.02
#>
#> 18 markers log-likelihood: -1170.482Finally, to check for alternative orders, use the ripple_seq function:
ripple_seq(LG2_f2_final, ws = 5, LOD = 3)The second argument, ws = 5, means that subsets (windows) of five markers will be permutated sequentially (5! orders for each window), to search for other plausible orders. The LOD argument means that only orders with LOD Score smaller than 3 will be printed.
The output shows sequences of five numbers, because ws = 5. They can be followed by an OK, if there are no alternative orders with LOD Scores smaller than LOD = 3 in absolute value, or by a list of alternative orders.
In this example, the first seven windows showed alternative orders with LOD smaller than LOD = 3. However, the best order was that obtained with the order_seq function (LOD = 0.00). If there were an alternative order more likely than the original, one should check the difference between them and, if necessary, change the order with (for example) functions drop_marker (see Section about using an arbitrary order) and try_seq, or simply by typing the new order. For that, use LG2_f2_final$seq.num to obtain the original order; then make the necessary changes (by copying and pasting) and use the function map to reestimate the genetic map for the new order.
The final estimated map can then be printed using
LG2_f2_final
#>
#> Printing map:
#>
#> Markers Position
#>
#> 2 M2 0.00
#> 10 M10 7.06
#> 41 M41 14.75
#> 11 M11 27.41
#> 43 M43 32.04
#> 32 M32 35.93
#> 45 M45 41.52
#> 54 M54 43.89
#> 66 M66 50.91
#> 61 M61 52.58
#> 25 M25 58.29
#> 5 M5 62.26
#> 51 M51 66.71
#> 91 SNP25 74.56
#> 33 M33 75.93
#> 90 SNP24 80.81
#> 37 M37 89.22
#> 8 M8 95.02
#>
#> 18 markers log-likelihood: -1170.482This is based on multipoint estimation of recombination fractions, using hidden Markov models.
Let us analyze linkage group 1. Extract markers from object LGs:
LG1_f2 <- make_seq(LGs_f2, 1)Construct the linkage map, by automatic usage of the try algorithm:
LG1_f2_ord <- order_seq(input.seq = LG1_f2, n.init = 5,
subset.search = "twopt",
twopt.alg = "rcd", THRES = 3,
touchdown = TRUE)The second round of try_seq added markers 9, 12, 29, 35, 58, 71 and 72 (try it; results not shown).
Get the order with all markers:
(LG1_f2_final <- make_seq(LG1_f2_ord, "force"))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 55 M55 0.00
#> 27 M27 7.69
#> 9 M9 10.16
#> 3 M3 21.27
#> 4 M4 23.59
#> 42 M42 30.05
#> 53 M53 31.58
#> 46 M46 35.89
#> 1 M1 44.68
#> 69 SNP3 50.50
#> 30 M30 53.66
#> 70 SNP4 56.30
#> 71 SNP5 69.40
#> 7 M7 71.82
#> 73 SNP7 73.59
#> 72 SNP6 78.44
#> 6 M6 87.45
#> 13 M13 92.62
#> 35 M35 97.08
#> 58 M58 104.80
#> 75 SNP9 115.63
#> 74 SNP8 121.94
#> 12 M12 123.60
#> 76 SNP10 125.48
#> 17 M17 131.02
#> 31 M31 134.53
#> 34 M34 143.64
#> 63 M63 143.64
#> 26 M26 151.82
#> 40 M40 155.42
#> 36 M36 163.42
#> 67 SNP1 166.30
#> 44 M44 168.84
#> 68 SNP2 171.48
#> 29 M29 180.72
#> 60 M60 184.04
#> 23 M23 192.44
#>
#> 37 markers log-likelihood: -2307.371Check the final map (results not shown):
ripple_seq(LG1_f2_final, ws = 5)No alternative order was found (please, try it).
Print it:
LG1_f2_final
#>
#> Printing map:
#>
#> Markers Position
#>
#> 55 M55 0.00
#> 27 M27 7.69
#> 9 M9 10.16
#> 3 M3 21.27
#> 4 M4 23.59
#> 42 M42 30.05
#> 53 M53 31.58
#> 46 M46 35.89
#> 1 M1 44.68
#> 69 SNP3 50.50
#> 30 M30 53.66
#> 70 SNP4 56.30
#> 71 SNP5 69.40
#> 7 M7 71.82
#> 73 SNP7 73.59
#> 72 SNP6 78.44
#> 6 M6 87.45
#> 13 M13 92.62
#> 35 M35 97.08
#> 58 M58 104.80
#> 75 SNP9 115.63
#> 74 SNP8 121.94
#> 12 M12 123.60
#> 76 SNP10 125.48
#> 17 M17 131.02
#> 31 M31 134.53
#> 34 M34 143.64
#> 63 M63 143.64
#> 26 M26 151.82
#> 40 M40 155.42
#> 36 M36 163.42
#> 67 SNP1 166.30
#> 44 M44 168.84
#> 68 SNP2 171.48
#> 29 M29 180.72
#> 60 M60 184.04
#> 23 M23 192.44
#>
#> 37 markers log-likelihood: -2307.371This is the final version of the map for this linkage group.
Extract markers from object LGs_f2:
LG3_f2 <- make_seq(LGs_f2, 3)Construct the linkage map, by automatic usage of try algorithm.
LG3_f2_ord <- order_seq(input.seq = LG3_f2, n.init = 5,
subset.search = "twopt",
twopt.alg = "rcd", THRES = 3,
touchdown = TRUE)We can see that in the second round of try_seq markers 50, 64, 78 and 86 were added (please, try it). A careful examination of the graphics can be a good source of information about how markers where placed.
Now, get the order with all markers:
(LG3_f2_final <- make_seq(LG3_f2_ord, "force"))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.56
#> 39 M39 8.94
#> 38 M38 15.81
#> 49 M49 23.71
#> 59 M59 24.88
#> 79 SNP13 35.91
#> 28 M28 39.40
#> 77 SNP11 41.34
#> 81 SNP15 45.45
#> 86 SNP20 55.98
#> 84 SNP18 61.10
#> 14 M14 62.67
#> 83 SNP17 64.90
#> 16 M16 70.87
#> 80 SNP14 80.95
#> 65 M65 86.12
#> 78 SNP12 91.55
#> 62 M62 101.24
#> 15 M15 104.96
#> 21 M21 108.67
#> 24 M24 112.05
#> 20 M20 118.55
#> 64 M64 119.07
#> 82 SNP16 128.75
#> 85 SNP19 134.28
#> 52 M52 136.53
#> 48 M48 141.80
#> 57 M57 147.83
#> 22 M22 151.20
#> 87 SNP21 158.16
#> 89 SNP23 163.24
#> 18 M18 164.72
#> 88 SNP22 167.97
#> 56 M56 179.82
#> 50 M50 182.43
#>
#> 36 markers log-likelihood: -2400.784Check the final map (not shown):
ripple_seq(LG3_f2_final, ws = 5)No better alternative order was observed.
Print it:
LG3_f2_final
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.56
#> 39 M39 8.94
#> 38 M38 15.81
#> 49 M49 23.71
#> 59 M59 24.88
#> 79 SNP13 35.91
#> 28 M28 39.40
#> 77 SNP11 41.34
#> 81 SNP15 45.45
#> 86 SNP20 55.98
#> 84 SNP18 61.10
#> 14 M14 62.67
#> 83 SNP17 64.90
#> 16 M16 70.87
#> 80 SNP14 80.95
#> 65 M65 86.12
#> 78 SNP12 91.55
#> 62 M62 101.24
#> 15 M15 104.96
#> 21 M21 108.67
#> 24 M24 112.05
#> 20 M20 118.55
#> 64 M64 119.07
#> 82 SNP16 128.75
#> 85 SNP19 134.28
#> 52 M52 136.53
#> 48 M48 141.80
#> 57 M57 147.83
#> 22 M22 151.20
#> 87 SNP21 158.16
#> 89 SNP23 163.24
#> 18 M18 164.72
#> 88 SNP22 167.97
#> 56 M56 179.82
#> 50 M50 182.43
#>
#> 36 markers log-likelihood: -2400.784It is possible to plot the recombination fraction matrix and LOD Scores based on a color scale using the function rf_graph_table. This matrix can be useful to make some diagnostics about the map.
Just as an example, let us place M38 at the end of linkage group 3 (wrong position) to check how we can see that in the graphic. First, let us to remove it from the linkage group:
temp_seq <- drop_marker(LG3_f2_final, 38)Now, we will force it to be positioned at the end of the linkage group:
(temp_seq <- add_marker(temp_seq, 38))
#>
#> Number of markers: 36
#> Markers in the sequence:
#> M47 M19 M39 M49 M59 SNP13 M28 SNP11 SNP15 SNP20 SNP18 M14 SNP17 M16 SNP14
#> M65 SNP12 M62 M15 M21 M24 M20 M64 SNP16 SNP19 M52 M48 M57 M22 SNP21 SNP23
#> M18 SNP22 M56 M50 M38
#>
#> Parameters not estimated.
(LG3_f2_wrong <- map(temp_seq))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.44
#> 39 M39 9.14
#> 49 M49 22.83
#> 59 M59 25.17
#> 79 SNP13 35.49
#> 28 M28 39.01
#> 77 SNP11 40.95
#> 81 SNP15 45.06
#> 86 SNP20 55.58
#> 84 SNP18 60.69
#> 14 M14 62.27
#> 83 SNP17 64.49
#> 16 M16 70.47
#> 80 SNP14 80.54
#> 65 M65 85.71
#> 78 SNP12 91.14
#> 62 M62 100.84
#> 15 M15 104.55
#> 21 M21 108.27
#> 24 M24 111.65
#> 20 M20 118.15
#> 64 M64 118.66
#> 82 SNP16 128.34
#> 85 SNP19 133.87
#> 52 M52 136.13
#> 48 M48 141.39
#> 57 M57 147.42
#> 22 M22 150.80
#> 87 SNP21 157.76
#> 89 SNP23 162.83
#> 18 M18 164.32
#> 88 SNP22 167.56
#> 56 M56 179.42
#> 50 M50 181.94
#> 38 M38 239.62
#>
#> 36 markers log-likelihood: -2527.534Examining the results, we can see there is a big gap at the end of linkage group 3 (between markers M50 and M38, as expected.
Let us now plot the recombination fraction matrix:
rf_graph_table(LG3_f2_wrong, inter = FALSE)The recombination fractions are plotted below the diagonal and the LOD Scores are plotted above the diagonal. The color scale varies from red (small distances and big LODs) to dark blue.
You can also plot an interactive version of this graphic, changing option inter to TRUE (the default):
rf_graph_table(LG3_f2_wrong)An interactive version of the graphic will pop up (not shown here). Clicking on the cell corresponding to two markers, you can see some useful information about them. For example, clicking on the cell corresponding to markers M47 and M19 you can see their names, types (co-dominant and dominant), recombination fraction (rf = 0.074) and LOD Score (LOD = 23). Clicking on a cell on the diagonal, some information about the corresponding marker is shown, including percentage of missing data. We found this to be very useful when evaluating the results of the ordering algorithms.
We clearly see a different pattern for marker M38. The blue cell, corresponding to markers M50 and M38, indicates a big recombination fraction between these markers as seen before (by clicking, rf = 0.4049). Moreover, we can see a group of red cells corresponding to marker M38 and other markers as M47, M19, M39 and M49. This pattern indicates small recombination fractions between marker M38 and these other markers. Hence, M38 is supposed to be close to them on the map.
Because we have enough evidence that marker M38 is misplaced, let us drop this marker and try to map it using the function try_seq:
temp_seq <- drop_marker(LG3_f2_wrong, 38)
temp_map <- map(temp_seq)
(temp_try <- try_seq(temp_map, 38))
#> 38 --> M38 : ....................................
#>
#> LOD scores correspond to the best linkage phase combination
#> for each position
#>
#> The symbol "*" outside the box indicates that more than one
#> linkage phase is possible for the corresponding position
#>
#>
#> Marker tested: 38
#>
#> Markers LOD
#> ======================
#> | |
#> | -27.70 | 1
#> | 47 |
#> | -24.58 | 2
#> | 19 |
#> | -9.72 | 3
#> | 39 |
#> | 0.00 | 4
#> | 49 |
#> | -17.43 | 5
#> | 59 |
#> | -27.50 | 6
#> | 79 |
#> | -68.03 | 7
#> | 28 |
#> | -72.63 | 8
#> | 77 |
#> | -68.45 | 9
#> | 81 |
#> | -53.97 | 10
#> | 86 |
#> | -69.71 | 11
#> | 84 |
#> | -82.70 | 12
#> | 14 |
#> | -79.77 | 13
#> | 83 |
#> | -69.78 | 14
#> | 16 |
#> | -71.51 | 15
#> | 80 |
#> | -92.91 | 16
#> | 65 |
#> | -93.17 | 17
#> | 78 |
#> | -80.94 | 18
#> | 62 |
#> | -96.11 | 19
#> | 15 |
#> | -97.12 | 20
#> | 21 |
#> | -98.70 | 21
#> | 24 |
#> | -94.17 | 22
#> | 20 |
#> | -89.13 | 23
#> | 64 |
#> | -87.60 | 24
#> | 82 |
#> | -98.60 | 25
#> | 85 |
#> | -110.51 | 26
#> | 52 |
#> | -97.67 | 27
#> | 48 |
#> | -100.08 | 28
#> | 57 |
#> | -109.84 | 29
#> | 22 |
#> | -95.03 | 30
#> | 87 |
#> | -104.05 | 31
#> | 89 |
#> | -120.57 | 32
#> | 18 |
#> | -110.57 | 33
#> | 88 |
#> | -84.52 | 34
#> | 56 |
#> | -87.05 | 35
#> | 50 |
#> | -71.32 | 36
#> | |
#> ======================We can see that the most likely position for marker M38 is between markers M39 and M49 (position 4). The patterns on the color matrix are now better. Therefore, you should do
(LG3_f2_final <- make_seq(temp_try, 4))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.73
#> 39 M39 7.98
#> 38 M38 15.25
#> 49 M49 23.40
#> 59 M59 23.59
#> 79 SNP13 35.25
#> 28 M28 38.84
#> 77 SNP11 40.76
#> 81 SNP15 44.87
#> 86 SNP20 55.63
#> 84 SNP18 61.02
#> 14 M14 62.77
#> 83 SNP17 65.38
#> 16 M16 71.87
#> 80 SNP14 82.29
#> 65 M65 87.62
#> 78 SNP12 93.25
#> 62 M62 103.57
#> 15 M15 107.44
#> 21 M21 111.33
#> 24 M24 114.57
#> 20 M20 121.30
#> 64 M64 121.47
#> 82 SNP16 131.43
#> 85 SNP19 137.55
#> 52 M52 140.24
#> 48 M48 145.48
#> 57 M57 151.43
#> 22 M22 154.83
#> 87 SNP21 161.82
#> 89 SNP23 166.93
#> 18 M18 168.41
#> 88 SNP22 171.68
#> 56 M56 183.61
#> 50 M50 186.22
#>
#> 36 markers log-likelihood: -2401.428We can also check the heatmap matrix for the other groups ordered. This is a good map diagnostic and can be used during the map building procedure:
rf_graph_table(LG1_f2_final)rf_graph_table(LG2_f2_final)In our example, we have reference genome chromosome and position informations for some of the markers, here we will examplify one method of using this informations to help build the genetic map.
With the CHROM informations in the input file, you can identify markers belonging to some chromosome using the function make_seq with the rf_2pts object. For example, assign the string "1" for the second argument to get chromosome 1 makers. The output sequence will be automatically ordered by POS informations.
CHR1 <- make_seq(twopts_f2, "1")
CHR1
#>
#> Number of markers: 10
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10
#>
#> Parameters not estimated.
CHR2 <- make_seq(twopts_f2, "2")
CHR3 <- make_seq(twopts_f2, "3")According to CHROM informations we have three defined linkage groups, now we can try to group the markers without chromossome informations to them using recombination informations. For this, we can use the function group_seq:
CHR_mks <- group_seq(input.2pts = twopts_f2, seqs = "CHROM", unlink.mks = mark_all_f2,
rm.repeated = TRUE)
#> Selecting markers:
#> group 1
#> .....................................
#> group 2
#> ................
#> group 3
#> .......................
#> Selecting markers:
#> group 1
#> ....................................
#> group 2
#> ...........................
#> group 3
#> ................
#> Selecting markers:
#> group 1
#> ..................
#> group 2
#> ...........................
#> group 3
#> .......................The function works as the function group, but considering preexisting sequences. Setting seqs argument with the string "CHROM", it will considered the preexisting sequences according to CHROM information. You can also indicate others preexisting sequences if it make sense for your study. For that, you should inform a list with objects of class sequences, as the example:
CHR_mks <- group_seq(input.2pts = twopts_f2, seqs = list(CHR1=CHR1, CHR2=CHR2, CHR3=CHR3),
unlink.mks = mark_all_f2, rm.repeated = TRUE)In this case, the command had the same effect of the previous, because we indicate chromossome sequences, but others sequences can be used.
The unlink.mks argument receive a object of class sequence, this define which markers will be tested to group with the sequences in seqs. In our example, we will indicate only the markers with no segregation distortion, using the sequence mark_no_dist. It is also possible to use the string "all" to test all the remining markers at the rf_2pts object.
In some cases, the same marker can group to more than one sequence, those markers will be considered repeated. We can choose if we want to remove or not (TRUE/FALSE) them of the output sequences, with the argument rm.repeated. Anyway, their numbers will be informed at the list repeateds in the output object. In the example case, there are no repeated markers. However, if they exists, it could indicate that their groups actually constitute the same group. Also, genotyping erros can generate repeated markers. Anyway, they deserves better investigations.
We can access detailed information about the results just printing:
CHR_mks
#> This is an object of class 'group_seq'
#> It was generated from the object "comb_example"
#>
#> Criteria used to assign markers to groups:
#> LOD = 3 , Maximum recombination fraction = 0.5
#>
#> No. markers in input sequences:
#> CHR1 : 10 markers
#> CHR2 : 13 markers
#> CHR3 : 2 markers
#>
#> No. unlinked input markers: 66 markers
#>
#> No. markers in output sequences:
#> CHR1 : 37 markers
#> CHR2 : 36 markers
#> CHR3 : 18 markers
#> No. unlinked: 0 markers
#> No. repeated: 0 markers
#>
#> Printing output sequences:
#> Group CHR1 : 37 markers
#>
#>
#> Group CHR2 : 36 markers
#>
#>
#> Group CHR3 : 18 markers
#>
#>
#> Unlinked markers: 0 markers
#>
#>
#> Repeated markers: 0 markers
#> Also, we can access the numbers of repeated markers with:
CHR_mks$repeated
#> [1] NAThe same way, we can access the output sequences:
CHR_mks$sequences$CHR1
#>
#> Number of markers: 37
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10 M1 M3 M4 M6 M7 M9 M12
#> M13 M17 M23 M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58
#> M60 M63
#>
#> Parameters not estimated.
# or
CHR_mks$sequences[[1]]
#>
#> Number of markers: 37
#> Markers in the sequence:
#> SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10 M1 M3 M4 M6 M7 M9 M12
#> M13 M17 M23 M26 M27 M29 M30 M31 M34 M35 M36 M40 M42 M44 M46 M53 M55 M58
#> M60 M63
#>
#> Parameters not estimated.For this function, optional arguments are LOD and max.rf, which define thresholds to be used when assigning markers to linkage groups. If none provided (default), criteria previously defined for the object rf_2pts are used.
Now we can order the markers in each group as we made before (Ordering markers within linkage groups). As showed, we can choose different approaches to order the markers.
To order those groups, first we will use order_seq function to access a preliminary order and after, we will edit some markers position or remove some of them according with they color pattern in rf_graph_table graphic, and other parameters as likelihood and map size.
CHR1_ord <- order_seq(CHR_mks$sequences$CHR1)
CHR1_frame <- make_seq(CHR1_ord, "force")rf_graph_table(CHR1_frame) # graphic not showedThe group is similar to that builded before with only recombinations informations. We will better explore differences in a latter step. Only marker 58 do not follow exactly the expected color pattern, then we will try to repositioned it.
CHR1_test_seq <- drop_marker(CHR1_frame, 58)
CHR1_test_map <- map(CHR1_test_seq)
CHR1_add58_seq <- try_seq(CHR1_test_map, 58)
#> 58 --> M58 : .....................................
CHR1_add58 <- make_seq(CHR1_add58_seq, 20) # marker 58 was placed at the same position as beforeComparing with and without marker 58:
CHR1_test_map
#>
#> Printing map:
#>
#> Markers Position
#>
#> 55 M55 0.00
#> 27 M27 7.69
#> 9 M9 10.16
#> 3 M3 21.27
#> 4 M4 23.59
#> 42 M42 30.05
#> 53 M53 31.58
#> 46 M46 35.89
#> 1 M1 44.80
#> 30 M30 49.37
#> 70 SNP4 51.38
#> 69 SNP3 56.20
#> 73 SNP7 72.37
#> 7 M7 75.61
#> 71 SNP5 77.23
#> 72 SNP6 82.56
#> 6 M6 91.08
#> 13 M13 96.09
#> 35 M35 103.34
#> 75 SNP9 118.14
#> 74 SNP8 124.60
#> 12 M12 126.31
#> 76 SNP10 128.21
#> 17 M17 133.85
#> 31 M31 137.36
#> 34 M34 146.47
#> 63 M63 146.47
#> 26 M26 154.64
#> 40 M40 158.25
#> 36 M36 166.25
#> 67 SNP1 169.12
#> 44 M44 171.66
#> 68 SNP2 174.31
#> 29 M29 183.54
#> 60 M60 186.87
#> 23 M23 195.27
#>
#> 36 markers log-likelihood: -2292.045rf_graph_table(CHR1_test_map) # graphic not showedCHR1_add58
#>
#> Printing map:
#>
#> Markers Position
#>
#> 55 M55 0.00
#> 27 M27 7.54
#> 9 M9 9.99
#> 3 M3 21.30
#> 4 M4 21.56
#> 42 M42 29.47
#> 53 M53 29.71
#> 46 M46 34.48
#> 1 M1 43.40
#> 30 M30 47.72
#> 70 SNP4 50.24
#> 69 SNP3 55.65
#> 73 SNP7 71.80
#> 7 M7 75.04
#> 71 SNP5 77.20
#> 72 SNP6 83.12
#> 6 M6 92.50
#> 13 M13 97.98
#> 35 M35 98.21
#> 58 M58 106.00
#> 75 SNP9 117.63
#> 74 SNP8 124.46
#> 12 M12 126.53
#> 76 SNP10 128.38
#> 17 M17 134.00
#> 31 M31 137.76
#> 34 M34 146.95
#> 63 M63 146.96
#> 26 M26 155.14
#> 40 M40 158.77
#> 36 M36 166.85
#> 67 SNP1 169.60
#> 44 M44 172.20
#> 68 SNP2 174.88
#> 29 M29 184.24
#> 60 M60 189.88
#> 23 M23 190.28
#>
#> 37 markers log-likelihood: -2327.558rf_graph_table(CHR1_add58)Based in those results, we decide to keep marker 58 in our map.
CHR1_final <- CHR1_add58Checking for better orders:
ripple_seq(CHR1_final)
#> 55-27-9-3-|-4-...
#> Alternative orders:
#> 55 27 9 3 ... : 0.00
#> 55 9 27 3 ... : -2.90
#>
#> ...-55-|-27-9-3-4-|-42-...
#> Alternative orders:
#> 27 9 3 4 ... : 0.00
#> 9 27 3 4 ... : -2.90
#>
#> ...-27-|-9-3-4-42-|-53-... OK
#>
#> ...-9-|-3-4-42-53-|-46-...
#> Alternative orders:
#> ... 3 4 42 53 ... : 0.00
#> ... 3 4 53 42 ... : -1.64
#> ... 3 53 4 42 ... : -1.87
#>
#> ...-3-|-4-42-53-46-|-1-...
#> Alternative orders:
#> ... 4 42 53 46 ... : 0.00
#> ... 4 53 42 46 ... : -1.64
#> ... 53 4 42 46 ... : -1.87
#>
#> ...-4-|-42-53-46-1-|-30-...
#> Alternative orders:
#> ... 42 53 46 1 ... : 0.00
#> ... 53 42 46 1 ... : -1.64
#>
#> ...-42-|-53-46-1-30-|-70-... OK
#>
#> ...-53-|-46-1-30-70-|-69-...
#> Alternative orders:
#> ... 46 1 30 70 ... : 0.00
#> ... 46 1 70 30 ... : -1.86
#>
#> ...-46-|-1-30-70-69-|-73-... OK
#>
#> ...-1-|-30-70-69-73-|-7-... OK
#>
#> ...-30-|-70-69-73-7-|-71-...
#> Alternative orders:
#> ... 70 69 73 7 ... : 0.00
#> ... 69 70 73 7 ... : -0.03
#>
#> ...-70-|-69-73-7-71-|-72-...
#> Alternative orders:
#> ... 69 73 7 71 ... : 0.00
#> ... 69 73 71 7 ... : -0.27
#> ... 69 71 7 73 ... : -0.27
#>
#> ...-69-|-73-7-71-72-|-6-...
#> Alternative orders:
#> ... 73 7 71 72 ... : 0.00
#> ... 72 73 7 71 ... : -0.18
#> ... 73 71 7 72 ... : -0.27
#> ... 71 7 73 72 ... : -0.27
#> ... 73 72 7 71 ... : -2.24
#> ... 72 71 7 73 ... : -2.38
#> ... 72 7 71 73 ... : -2.59
#> ... 72 73 71 7 ... : -2.64
#>
#> ...-73-|-7-71-72-6-|-13-...
#> Alternative orders:
#> ... 7 71 72 6 ... : 0.00
#> ... 71 7 72 6 ... : -0.27
#> ... 72 7 71 6 ... : -2.24
#>
#> ...-7-|-71-72-6-13-|-35-... OK
#>
#> ...-71-|-72-6-13-35-|-58-...
#> Alternative orders:
#> ... 72 6 13 35 ... : 0.00
#> ... 72 35 6 13 ... : -2.59
#> ... 72 6 35 13 ... : -2.74
#>
#> ...-72-|-6-13-35-58-|-75-...
#> Alternative orders:
#> ... 6 13 35 58 ... : 0.00
#> ... 35 6 13 58 ... : -2.59
#> ... 6 35 13 58 ... : -2.74
#>
#> ...-6-|-13-35-58-75-|-74-...
#> Alternative orders:
#> ... 13 35 58 75 ... : 0.00
#> ... 35 13 58 75 ... : -2.74
#>
#> ...-13-|-35-58-75-74-|-12-... OK
#>
#> ...-35-|-58-75-74-12-|-76-...
#> Alternative orders:
#> ... 58 75 74 12 ... : 0.00
#> ... 58 75 12 74 ... : -1.71
#>
#> ...-58-|-75-74-12-76-|-17-...
#> Alternative orders:
#> ... 75 74 12 76 ... : 0.00
#> ... 75 12 74 76 ... : -1.71
#> ... 75 76 12 74 ... : -2.23
#>
#> ...-75-|-74-12-76-17-|-31-...
#> Alternative orders:
#> ... 74 12 76 17 ... : 0.00
#> ... 12 74 76 17 ... : -1.71
#> ... 76 12 74 17 ... : -2.23
#>
#> ...-74-|-12-76-17-31-|-34-... OK
#>
#> ...-12-|-76-17-31-34-|-63-... OK
#>
#> ...-76-|-17-31-34-63-|-26-...
#> Alternative orders:
#> ... 17 31 34 63 ... : 0.00
#> ... 17 31 63 34 ... : -0.02
#>
#> ...-17-|-31-34-63-26-|-40-...
#> Alternative orders:
#> ... 31 34 63 26 ... : 0.00
#> ... 31 63 34 26 ... : -0.02
#>
#> ...-31-|-34-63-26-40-|-36-...
#> Alternative orders:
#> ... 34 63 26 40 ... : 0.00
#> ... 63 34 26 40 ... : -0.02
#>
#> ...-34-|-63-26-40-36-|-67-... OK
#>
#> ...-63-|-26-40-36-67-|-44-... OK
#>
#> ...-26-|-40-36-67-44-|-68-... OK
#>
#> ...-40-|-36-67-44-68-|-29-...
#> Alternative orders:
#> ... 36 68 44 67 ... : 0.00
#> ... 36 67 44 68 ... : -0.21
#>
#> ...-36-|-67-44-68-29-|-60-...
#> Alternative orders:
#> ... 68 44 67 29 ... : 0.00
#> ... 67 44 68 29 ... : -0.21
#>
#> ...-67-|-44-68-29-60-|-23-... OK
#>
#> 44-|-68-29-60-23
#> Alternative orders:
#> 68 29 60 23 ... : 0.00
#> 68 29 23 60 ... : -0.17CHR2_ord <- order_seq(CHR_mks$sequences$CHR2)
CHR2_frame <- make_seq(CHR2_ord, "force")rf_graph_table(CHR2_frame) # graphic not showedThe marker 20 presented a color pattern different of expected.
CHR2_test_seq <- drop_marker(CHR2_frame, 20)
CHR2_test_map <- map(CHR2_test_seq)
CHR2_add20_seq <- try_seq(CHR2_test_map, 20)
#> 20 --> M20 : ....................................
CHR2_add20 <- make_seq(CHR2_add20_seq, 20) # marker 20 was placed at the same position as beforeComparing with and without marker 58:
CHR2_test_map
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.56
#> 39 M39 8.94
#> 38 M38 15.81
#> 49 M49 23.70
#> 59 M59 24.88
#> 79 SNP13 35.91
#> 28 M28 39.40
#> 77 SNP11 41.34
#> 81 SNP15 45.45
#> 86 SNP20 55.98
#> 84 SNP18 61.10
#> 14 M14 62.67
#> 83 SNP17 64.90
#> 16 M16 70.87
#> 80 SNP14 80.95
#> 65 M65 86.12
#> 78 SNP12 91.55
#> 62 M62 101.25
#> 15 M15 104.96
#> 21 M21 108.67
#> 24 M24 112.12
#> 64 M64 118.31
#> 82 SNP16 127.51
#> 85 SNP19 133.23
#> 52 M52 135.56
#> 48 M48 140.97
#> 57 M57 146.98
#> 22 M22 150.36
#> 87 SNP21 157.20
#> 18 M18 160.62
#> 89 SNP23 162.17
#> 88 SNP22 167.09
#> 56 M56 178.98
#> 50 M50 181.59
#>
#> 35 markers log-likelihood: -2369.779rf_graph_table(CHR1_test_map)CHR2_add20
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 19 M19 7.73
#> 39 M39 7.98
#> 38 M38 15.25
#> 49 M49 23.40
#> 59 M59 23.59
#> 79 SNP13 35.25
#> 28 M28 38.84
#> 77 SNP11 40.76
#> 81 SNP15 44.87
#> 86 SNP20 55.63
#> 84 SNP18 61.02
#> 14 M14 62.77
#> 83 SNP17 65.38
#> 16 M16 71.88
#> 80 SNP14 82.35
#> 65 M65 87.70
#> 78 SNP12 93.38
#> 62 M62 104.30
#> 20 M20 117.45
#> 15 M15 130.08
#> 21 M21 133.59
#> 24 M24 138.74
#> 64 M64 139.02
#> 82 SNP16 150.70
#> 85 SNP19 156.94
#> 52 M52 159.69
#> 48 M48 165.01
#> 57 M57 170.97
#> 22 M22 174.37
#> 87 SNP21 181.28
#> 18 M18 184.76
#> 89 SNP23 186.30
#> 88 SNP22 191.21
#> 56 M56 203.16
#> 50 M50 205.77
#>
#> 36 markers log-likelihood: -2476.901rf_graph_table(CHR2_add20) # graphic not showedBased in those results, we decide to not include marker 20 in our map.
CHR2_final <- CHR2_test_mapChecking for better orders:
ripple_seq(CHR2_final)
#> 47-19-39-38-|-49-...
#> Alternative orders:
#> 47 19 39 38 ... : 0.00
#> 47 39 19 38 ... : -0.05
#>
#> ...-47-|-19-39-38-49-|-59-...
#> Alternative orders:
#> 19 39 38 49 ... : 0.00
#> 39 19 38 49 ... : -0.05
#>
#> ...-19-|-39-38-49-59-|-79-...
#> Alternative orders:
#> ... 39 38 49 59 ... : 0.00
#> ... 39 38 59 49 ... : -0.09
#>
#> ...-39-|-38-49-59-79-|-28-...
#> Alternative orders:
#> ... 38 49 59 79 ... : 0.00
#> ... 38 59 49 79 ... : -0.09
#>
#> ...-38-|-49-59-79-28-|-77-...
#> Alternative orders:
#> ... 49 59 79 28 ... : 0.00
#> ... 59 49 79 28 ... : -0.09
#>
#> ...-49-|-59-79-28-77-|-81-... OK
#>
#> ...-59-|-79-28-77-81-|-86-...
#> Alternative orders:
#> ... 81 77 28 79 ... : 0.00
#> ... 77 81 28 79 ... : -1.26
#>
#> ...-79-|-28-77-81-86-|-84-...
#> Alternative orders:
#> ... 28 77 81 86 ... : 0.00
#> ... 28 81 77 86 ... : -2.46
#>
#> ...-28-|-77-81-86-84-|-14-...
#> Alternative orders:
#> ... 77 81 86 84 ... : 0.00
#> ... 81 77 86 84 ... : -2.46
#>
#> ...-77-|-81-86-84-14-|-83-...
#> Alternative orders:
#> ... 81 86 84 14 ... : 0.00
#> ... 81 84 14 86 ... : -2.53
#>
#> ...-81-|-86-84-14-83-|-16-...
#> Alternative orders:
#> ... 86 84 14 83 ... : 0.00
#> ... 86 83 14 84 ... : -1.63
#> ... 84 14 86 83 ... : -2.53
#>
#> ...-86-|-84-14-83-16-|-80-...
#> Alternative orders:
#> ... 84 14 83 16 ... : 0.00
#> ... 83 14 84 16 ... : -1.63
#>
#> ...-84-|-14-83-16-80-|-65-... OK
#>
#> ...-14-|-83-16-80-65-|-78-...
#> Alternative orders:
#> ... 83 16 80 65 ... : 0.00
#> ... 83 16 65 80 ... : -2.19
#>
#> ...-83-|-16-80-65-78-|-62-...
#> Alternative orders:
#> ... 16 80 65 78 ... : 0.00
#> ... 16 65 80 78 ... : -2.19
#> ... 16 65 78 80 ... : -2.34
#> ... 16 80 78 65 ... : -2.50
#> ... 16 78 65 80 ... : -2.96
#>
#> ...-16-|-80-65-78-62-|-15-...
#> Alternative orders:
#> ... 80 65 78 62 ... : 0.00
#> ... 65 80 78 62 ... : -2.19
#> ... 65 78 80 62 ... : -2.34
#> ... 80 78 65 62 ... : -2.50
#> ... 78 65 80 62 ... : -2.96
#>
#> ...-80-|-65-78-62-15-|-21-...
#> Alternative orders:
#> ... 65 78 62 15 ... : 0.00
#> ... 78 65 62 15 ... : -2.50
#>
#> ...-65-|-78-62-15-21-|-24-... OK
#>
#> ...-78-|-62-15-21-24-|-64-... OK
#>
#> ...-62-|-15-21-24-64-|-82-...
#> Alternative orders:
#> ... 15 21 24 64 ... : 0.00
#> ... 15 21 64 24 ... : -0.38
#>
#> ...-15-|-21-24-64-82-|-85-...
#> Alternative orders:
#> ... 21 24 64 82 ... : 0.00
#> ... 21 64 24 82 ... : -0.38
#>
#> ...-21-|-24-64-82-85-|-52-...
#> Alternative orders:
#> ... 24 64 82 85 ... : 0.00
#> ... 64 24 82 85 ... : -0.38
#>
#> ...-24-|-64-82-85-52-|-48-...
#> Alternative orders:
#> ... 64 82 85 52 ... : 0.00
#> ... 64 52 85 82 ... : -1.54
#>
#> ...-64-|-82-85-52-48-|-57-...
#> Alternative orders:
#> ... 82 85 52 48 ... : 0.00
#> ... 52 85 82 48 ... : -1.54
#>
#> ...-82-|-85-52-48-57-|-22-... OK
#>
#> ...-85-|-52-48-57-22-|-87-... OK
#>
#> ...-52-|-48-57-22-87-|-18-... OK
#>
#> ...-48-|-57-22-87-18-|-89-... OK
#>
#> ...-57-|-22-87-18-89-|-88-...
#> Alternative orders:
#> ... 22 87 18 89 ... : 0.00
#> ... 22 87 89 18 ... : 0.00
#> ... 22 89 18 87 ... : -0.02
#> ... 22 18 89 87 ... : -0.78
#>
#> ...-22-|-87-18-89-88-|-56-...
#> Alternative orders:
#> ... 87 18 89 88 ... : 0.00
#> ... 87 89 18 88 ... : 0.00
#> ... 89 18 87 88 ... : -0.02
#> ... 87 88 18 89 ... : -0.61
#> ... 89 18 88 87 ... : -0.69
#> ... 18 89 87 88 ... : -0.78
#> ... 18 89 88 87 ... : -1.76
#>
#> ...-87-|-18-89-88-56-|-50-...
#> Alternative orders:
#> ... 18 89 88 56 : 0.00
#> ... 89 18 88 56 : 0.00
#> ... 88 18 89 56 : -0.61
#>
#> 18-|-89-88-56-50
#> Alternative orders:
#> 89 88 56 50 ... : 0.00
#> 89 88 50 56 ... : -2.27CHR3_ord <- order_seq(CHR_mks$sequences$CHR3)
CHR3_frame <- make_seq(CHR3_ord, "force")rf_graph_table(CHR3_frame) # graphic not showedHere, marker 32 have color pattern too different of the expected, removing it could influentiate in other markers ordering. Then, we will remove it and search for a new order.
CHR3_test_seq <- drop_marker(CHR3_frame, 32)
CHR3_test_ord <- order_seq(CHR3_test_seq)
CHR3_test_map <- make_seq(CHR3_test_ord, "force")The order did not change.
Trying to add the marker again.
CHR3_add32_seq <- try_seq(CHR3_test_map, 32)
CHR3_add32 <- make_seq(CHR3_add32_seq, 13) # Marker 32 keeped in the mapCHR3_final <- CHR3_add32
rf_graph_table(CHR3_final, inter = FALSE)Checking for better orders:
ripple_seq(CHR3_final)
#> 8-37-91-33-|-90-...
#> Alternative orders:
#> 8 37 91 33 ... : 0.00
#> 8 37 33 91 ... : -0.79
#>
#> ...-8-|-37-91-33-90-|-51-...
#> Alternative orders:
#> 37 90 33 91 ... : 0.00
#> 37 91 33 90 ... : -0.84
#> 37 90 91 33 ... : -1.04
#> 37 33 91 90 ... : -1.62
#>
#> ...-37-|-91-33-90-51-|-5-...
#> Alternative orders:
#> ... 90 33 91 51 ... : 0.00
#> ... 91 33 90 51 ... : -0.84
#> ... 90 91 33 51 ... : -1.04
#> ... 33 91 90 51 ... : -1.62
#>
#> ...-91-|-33-90-51-5-|-25-... OK
#>
#> ...-33-|-90-51-5-25-|-61-... OK
#>
#> ...-90-|-51-5-25-61-|-66-... OK
#>
#> ...-51-|-5-25-61-66-|-54-... OK
#>
#> ...-5-|-25-61-66-54-|-45-... OK
#>
#> ...-25-|-61-66-54-45-|-32-...
#> Alternative orders:
#> ... 61 66 54 45 ... : 0.00
#> ... 61 66 45 54 ... : -2.68
#>
#> ...-61-|-66-54-45-32-|-43-...
#> Alternative orders:
#> ... 66 54 45 32 ... : 0.00
#> ... 66 54 32 45 ... : -0.34
#> ... 66 45 54 32 ... : -2.68
#>
#> ...-66-|-54-45-32-43-|-11-...
#> Alternative orders:
#> ... 54 45 32 43 ... : 0.00
#> ... 54 32 45 43 ... : -0.34
#> ... 54 45 43 32 ... : -1.94
#> ... 45 54 32 43 ... : -2.68
#>
#> ...-54-|-45-32-43-11-|-2-...
#> Alternative orders:
#> ... 45 32 43 11 ... : 0.00
#> ... 32 45 43 11 ... : -0.34
#> ... 45 43 32 11 ... : -1.94
#> ... 45 43 11 32 ... : -2.68
#>
#> ...-45-|-32-43-11-2-|-10-...
#> Alternative orders:
#> ... 32 43 11 2 ... : 0.00
#> ... 32 43 2 11 ... : -0.73
#> ... 43 32 11 2 ... : -1.94
#> ... 43 11 32 2 ... : -2.68
#> ... 43 11 2 32 ... : -2.75
#> ... 43 32 2 11 ... : -2.78
#>
#> ...-32-|-43-11-2-10-|-41-...
#> Alternative orders:
#> ... 43 11 2 10 : 0.00
#> ... 43 2 11 10 : -0.73
#>
#> 43-|-11-2-10-41
#> Alternative orders:
#> 11 2 10 41 ... : 0.00
#> 2 11 10 41 ... : -0.73If you have some information about the order of the markers, for example, from a reference genome or previously published paper, you can define a sequence of those markers in a specific order (using the function make_seq) and then use the function map to estimate the final genetic map (based on multipoint analysis). For example, let us assume that we know that the following markers are ordered in this sequence: M47, M38, M59, M16, M62, M21, M20, M48 and M22. In this case, select them from the two point analysis, and use function map:
LG3seq_f2 <- make_seq(twopts_f2, c(47, 38, 59, 16, 62, 21, 20, 48, 22))
(LG3seq_f2_map <- map(LG3seq_f2))
#>
#> Printing map:
#>
#> Markers Position
#>
#> 47 M47 0.00
#> 38 M38 14.49
#> 59 M59 23.99
#> 16 M16 40.13
#> 62 M62 50.60
#> 21 M21 57.84
#> 20 M20 67.16
#> 48 M48 75.76
#> 22 M22 84.62
#>
#> 9 markers log-likelihood: -947.1229To see the correspondence between marker names and numbers, use function marker_type:
marker_type(LG3seq_f2_map)
#> Marker 47 ( M47 ) --> AA : AB : BB (1:2:1)
#> Marker 38 ( M38 ) --> AA : AB : BB (1:2:1)
#> Marker 59 ( M59 ) --> Not AA : AA (3:1)
#> Marker 16 ( M16 ) --> AA : AB : BB (1:2:1)
#> Marker 62 ( M62 ) --> AA : AB : BB (1:2:1)
#> Marker 21 ( M21 ) --> AA : AB : BB (1:2:1)
#> Marker 20 ( M20 ) --> Not BB : BB (3:1)
#> Marker 48 ( M48 ) --> AA : AB : BB (1:2:1)
#> Marker 22 ( M22 ) --> AA : AB : BB (1:2:1)If one needs to add or drop markers from a predefined sequence, functions add_marker and drop_marker can be used. For example, to add markers M18, M56 and M50 at the end of LG3seq_f2_map:
(LG3seq_f2_map <- add_marker(LG3seq_f2_map, c(18, 56, 50)))
#>
#> Number of markers: 12
#> Markers in the sequence:
#> M47 M38 M59 M16 M62 M21 M20 M48 M22 M18 M56 M50
#>
#> Parameters not estimated.Removing markers M59 and 21 from LG3seq_f2_map:
(LG3seq_f2_map <- drop_marker(LG3seq_f2_map, c(59, 21)))
#>
#> Number of markers: 10
#> Markers in the sequence:
#> M47 M38 M16 M62 M20 M48 M22 M18 M56 M50
#>
#> Parameters not estimated.We can draw a genetic map for all linkage groups using the function draw_map. First we have to create a list of ordered linkage groups:
maps_list <- list(LG1_f2_final, LG2_f2_final, LG3_f2_final)Then use function draw_map for this list:
draw_map(maps_list, names = TRUE, grid = TRUE, cex.mrk = 0.7)We also can draw a map for a specific linkage group:
draw_map(LG1_f2_final, names = TRUE, grid = TRUE, cex.mrk = 0.7)Let’s use this function to overview the maps builded by different strategies in this tutorial:
map_list_all <- list(LG1_f2_final, CHR1_final, LG2_f2_final, CHR3_final, CHR2_final, LG3_f2_final)
draw_map(map_list_all, names = TRUE, grid = TRUE, cex.mrk = 0.7)Function draw_map draws a very simple graphic representation of the genetic map. However, once the distances are estimated, better map figures can be drawn by the user with any appropriate software. There are several free software that can be used, such as MapChart (Voorrips, 2002).
At this point it should be clear that any potential OneMap user must have some knowledge about genetic mapping and also the R language, because the analysis is not done with only one mouse click. In the future, perhaps a graphical interface will be made available to make this software a lot easier to use.
We do hope that OneMap is useful to researchers interested in genetic mapping in outcrossing or inbred-based populations. Any suggestions and critics are welcome.
sessionInfo()
#> R version 3.4.2 (2017-09-28)
#> Platform: x86_64-pc-linux-gnu (64-bit)
#> Running under: Ubuntu 16.04.3 LTS
#>
#> Matrix products: default
#> BLAS: /usr/lib/libblas/libblas.so.3.6.0
#> LAPACK: /usr/lib/lapack/liblapack.so.3.6.0
#>
#> locale:
#> [1] LC_CTYPE=pt_BR.UTF-8 LC_NUMERIC=C
#> [3] LC_TIME=pt_BR.UTF-8 LC_COLLATE=C
#> [5] LC_MONETARY=pt_BR.UTF-8 LC_MESSAGES=en_US.UTF-8
#> [7] LC_PAPER=pt_BR.UTF-8 LC_NAME=C
#> [9] LC_ADDRESS=C LC_TELEPHONE=C
#> [11] LC_MEASUREMENT=pt_BR.UTF-8 LC_IDENTIFICATION=C
#>
#> attached base packages:
#> [1] stats graphics grDevices utils datasets methods base
#>
#> other attached packages:
#> [1] onemap_2.1.0 rmarkdown_1.6 knitr_1.17
#>
#> loaded via a namespace (and not attached):
#> [1] Rcpp_0.12.13 magrittr_1.5 maps_3.2.0 munsell_0.4.3
#> [5] colorspace_1.3-2 rlang_0.1.2 stringr_1.2.0 plyr_1.8.4
#> [9] tcltk_3.4.2 fields_9.0 tools_3.4.2 dotCall64_0.9-04
#> [13] grid_3.4.2 tkrplot_0.0-23 spam_2.1-1 gtable_0.2.0
#> [17] htmltools_0.3.6 yaml_2.1.14 lazyeval_0.2.0 rprojroot_1.2
#> [21] digest_0.6.12 tibble_1.3.4 reshape2_1.4.2 ggplot2_2.2.1
#> [25] codetools_0.2-15 evaluate_0.10.1 labeling_0.3 stringi_1.1.5
#> [29] compiler_3.4.2 scales_0.5.0 backports_1.1.1Adler, J. R in a Nutshell A Desktop Quick Reference, 2009.
Broman, K. W., Wu, H., Churchill, G., Sen, S., Yandell, B. qtl: Tools for analyzing QTL experiments R package version 1.09-43, 2008.
Buetow, K. H., Chakravarti, A. Multipoint gene mapping using seriation. I. General methods. American Journal of Human Genetics 41, 180-188, 1987.
Doerge, R.W. Constructing genetic maps by rapid chain delineation. Journal of Agricultural Genomics 2, 1996.
Garcia, A.A.F., Kido, E.A., Meza, A.N., Souza, H.M.B., Pinto, L.R., Pastina, M.M., Leite, C.S., Silva, J.A.G., Ulian, E.C., Figueira, A. and Souza, A.P. Development of an integrated genetic map of a sugarcane Saccharum spp. commercial cross, based on a maximum-likelihood approach for estimation of linkage and linkage phases. Theoretical and Applied Genetics 112, 298-314, 2006.
Haldane, J. B. S. The combination of linkage values and the calculation of distance between the loci of linked factors. Journal of Genetics 8, 299-309, 1919.
Jiang, C. and Zeng, Z.-B. Mapping quantitative trait loci with dominant and missing markers in various crosses from two inbred lines. Genetica 101, 47-58, 1997.
Kosambi, D. D. The estimation of map distance from recombination values. Annuaire of Eugenetics 12, 172-175, 1944.
Lander, E. S. and Green, P. Construction of multilocus genetic linkage maps in humans. Proc. Natl. Acad. Sci. USA 84, 2363-2367, 1987.
Lander, E.S., Green, P., Abrahanson, J., Barlow, A., Daly, M.J., Lincoln, S.E. and Newburg, L. MAPMAKER, An interactive computing package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174-181, 1987.
Lincoln, S. E., Daly, M. J. and Lander, E. S. Constructing genetic linkage maps with MAPMAKER/EXP Version 3.0: a tutorial and reference manual. A Whitehead Institute for Biomedical Research Technical Report 1993.
Matloff, N. The Art of R Programming. 2011. 1st ed. San Francisco, CA: No Starch Press, Inc., 404 pages.
Margarido, G. R. A., Souza, A.P. and Garcia, A. A. F. OneMap: software for genetic mapping in outcrossing species. Hereditas 144, 78-79, 2007.
Mollinari, M., Margarido, G. R. A., Vencovsky, R. and Garcia, A. A. F. Evaluation of algorithms used to order markers on genetics maps. Heredity 103, 494-502, 2009.
Oliveira, K.M., Pinto, L.R., Marconi, T.G., Margarido, G.R.A., Pastina, M.M., Teixeira, L.H.M., Figueira, A.M., Ulian, E.C., Garcia, A.A.F., Souza, A.P. Functional genetic linkage map on EST-markers for a sugarcane (Saccharum spp.) commercial cross. Molecular Breeding 20, 189-208, 2007.
Oliveira, E. J., Vieira, M. L. C., Garcia, A. A. F., Munhoz, C. F.,Margarido, G. R.A., Consoli, L., Matta, F. P., Moraes, M. C., Zucchi, M. I., and Fungaro,M. H. P. An Integrated Molecular Map of Yellow Passion Fruit Based on Simultaneous Maximum-likelihood Estimation of Linkage and Linkage Phases J. Amer. Soc. Hort. Sci. 133, 35-41, 2008.
Tan, Y., Fu, Y. A novel method for estimating linkage maps. Genetics 173, 2383-2390, 2006.
Van Os H, Stam P, Visser R.G.F., Van Eck H.J. RECORD: a novel method for ordering loci on a genetic linkage map. Theor Appl Genet 112, 30-40, 2005.
Voorrips, R.E. MapChart: software for the graphical presentation of linkage maps and QTLs. Journal of Heredity 93, 77-78, 2002.
Wang S., Basten, C. J. and Zeng Z.-B. Windows QTL Cartographer 2.5. Department of Statistics, North Carolina State University, Raleigh, NC, 2010. http://statgen.ncsu.edu/qtlcart/WQTLCart.htm
Wu, R., Ma, C.X., Painter, I. and Zeng, Z.-B. Simultaneous maximum likelihood estimation of linkage and linkage phases in outcrossing species. Theoretical Population Biology 61, 349-363, 2002a.
Wu, R., Ma, C.-X., Wu, S. S. and Zeng, Z.-B. Linkage mapping of sex-specific differences. Genetical Research 79, 85-96, 2002b.