Lep-MAP3 Activity
Brought to you by:
lep-map
Activity for Lep-MAP3
-
Lorenzo posted a comment on discussion General Discussion
Hello Pasi, Replying here for completeness and for future users. I have now encountered this pattern twice. Once with the multifamily design described above, and once with single families of 2 parents x ~100 F1 offspring. For the multifamily map we were able to fix this by using the hyperPhaser parameter and increasing the number of iterations the OrderMarkers module was run for. For the single family maps we tried everything in OrderMarkers, but nothing fixed it, so we are now using FitSteFunction...
-
Hello Pasi, Thank you very much for your reply. I will attempt to expand on your three suggested options, to ensure that we understand them correctly, and for future reference for other users: Three ways to fix this: 1) evaluate the map in the physical order (evaluateOrder=phys_order.txt, improveOrder=0) The rationale for this option is to constrain the order (1st, 2nd, 3rd, etc) of markers to their physical position on the assembly during the OrderMarkers2 step, to ensure a monotonically increasing...
-
Samantha Howitt posted a comment on discussion General Discussion
Hi Pasi, Thank you very much for your suggestions and explanation, we will be sure to read and try lep-anchor. Cheers, Sam
-
Ana Kurdadze posted a comment on discussion General Discussion
Hi Pasi, I found this solution in the discussions, and it appears to resolve the issue. zcat $PAT_CELL_VCF|$JAVA -cp $BIN ParentCall2 \ data=$PEDIGREE \ vcfFile=- \ removeNonInformative=1 > $OUT_DATA Thanks !
-
Hi Pasi, Thanks a lot for the quick response! After reading some discussions, I tried the following: Add PATERNAL_PAT as grandparent, variant sites 1/1 Add DUMMY_MAT as dummy_gp, always 0/0 Add DUMMY_F1_PAT as offspring, 0/1 at every site where PATERNAL_PAT is 1/1 Add DUMMY_F1_MAT always 0/0 This the command I run: $JAVA -cp $BIN ParentCall2 \ data=$PEDIGREE \ vcfFile=$PAT_CELL_VCF \ removeNonInformative=1 > $OUT_DATA After multiple attempts, I still get the following error Warning: Different number...
-
Hi Pasi, Thanks a lot for the quick response! After reading some discussions, I tried the following: Add PATERNAL_PAT as grandparent, variant sites 1/1 Add DUMMY_MAT as dummy_gp, always 0/0 Add DUMMY_F1_PAT as offspring, 0/1 at every site where PATERNAL_PAT is 1/1 Add DUMMY_F1_MAT always 0/0 This the command I run: $JAVA -cp $BIN ParentCall2 \ data=$PEDIGREE \ vcfFile=$PAT_CELL_VCF \ removeNonInformative=1 > $OUT_DATA ``` I still get the following error Warning: Different number of grandparents (4...
-
Pasi Rastas posted a comment on discussion General Discussion
Dear Sam, Thank you for your question. The maps look about as good as they can, being build "de novo" without genome coordinates. My gut feeling is that the double lines are caused because some markers have missing/erroneous genotypes. This uncertainty of map positions can be obtained from the map intervals (parameter calculateIntervals=file), I did discuss this in detail in my Lep-Anchor paper. Basically I have suggested three ways to fix this: 1) evaluate the map in the physical order (evaluateOrder=phys_order.txt,...
-
Dear Ana, Thank you for your question. Please search the discussions with word "haploid", it seemed to find many relevant topics that might help you (sorry for my lazy answer). Lep-MAP3 can analyse you data, but you have to "trick" Lep-MAP3 to think the data is in one of its supported formats F1/F2/selfing cross. Anyway, I checked you pedigree and you should add two parents to your family (real parent and dummy, or grandparent(s) and dummy parents). Now the other parent is 0 that does not pass Lep-MAP3....
-
Dear Pasi, Thank you for the software and all your comments on this forum, they have been very helpful for us on our linkage mapping project. Overall, our mapping has been going really well, however we are having some trouble fine-tuning order markers, and observe regions where the genetic position does not converge, so we see 2 -4 lines in some regions when comparing to the physical maps. We have tried a variety of things to remove these (detailed below), and whilst they improve on some LGs, they...
-
Hi Pasi, Thanks a lot for the software! I have scRna-seq data from over 26000 pollen cells from F1 sample. I also have variant sites from alignment of the paternal reads to the maternal reference genome. I am running ParentCall2 in halfSibs=1 mode. I have attached the pedigree. I'm getting error java.lang.NumberFormatException: For input string: "PARENT_PAT" at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67) at java.base/java.lang.Integer.parseInt(Integer.java:565)...
-
Hi Pasti, Thanks a lot for the software! I have scRna-seq data from over 26000 pollen cells from F1 sample. I also have variant sites from alignment of the paternal reads to the maternal reference genome. I am running ParentCall2 in halfSibs=1 mode. I have attached the pedigree. I'm getting error java.lang.NumberFormatException: For input string: "PARENT_PAT" at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67) at java.base/java.lang.Integer.parseInt(Integer.java:565)...
-
Dear Huan, Thank you for your question. Getting markers into right number of groups can be difficult. There are many discussion threads that might help here. First of all, always make sure that your family structure is correct! Then make sure to use distortionLod=1 (or filter distorted markers in other ways). This should make the marker assignment more robust. With 400 individuals in one family, you could use a lodLimit of about 40. And you can always use both families to assign markers as well if...
-
hfan posted a comment on discussion General Discussion
Hi Pasi, I am working on two families. For family one, I got very nice results at the SeparateChromosomes2 stage. I decided on which lodLimit to use by the lowest number where the first group (1) mainly goes to one chromosome, and I get the same number of major LGs as the number of chromosomes in my genome, with a one-on-one relationship. However, I cannot get such glear pattern for family two. If I use the same rule for lodLimit, I end up with not enough number of LGs. Then if I continue to run...
-
Dear Huan, The squeeze removes map positions from the map ends if there are no support for them, basically making the map start at 0cM. Cheers, Pasi
-
Hi Pasi, The FitStepFunction worked really well! Thank you! I was reinventing the wheel by doing some iterative loess smoothing... One quick question, what is the last line of FitStepFunction's printscreen saying about Squeeze map? e.g. here is from one of my chromosome: java -cp ~/build/lep-anchor-code/bin FitStepFunction map=order.txt autodetecting noChromosome=1 The most abundant contig is chr1 with 54888 markers +orientation 37811 -orientation 3443 Using + orientation Squeeze map 0->57 1377->1377...
-
Hi Pasi, The FitStepFunction worked really well! Thank you! I was reinventing the wheel by doing some iterative loess smoothing... One quick question, what is the last line of FitStepFunction's printscreen saying about Squeeze map? e.g. here is from one of my chromosome: java -cp ~/build/lep-anchor-code/bin FitStepFunction map=order.txt autodetecting noChromosome=1 The most abundant contig is chr1 with 54888 markers +orientation 37811 -orientation 3443 Using + orientation Squeeze map 0->57 1377->1377...
-
Dear Huan, Thank you for your question. Having this kind of map data is normal. My understanding is that there are transposable elements and other repeats that make some markers map into wrong chromosomes and places. This can be fixed by removing markers in the wrong chromosomes. I don't know how much such markers affect other analysis, but these stand out in linkage maps. Addition to these "jumping" markers, within chromosomes the map data is not necessary informative enough to put all markers into...
-
Hi Pasi and dear users, I wanted to phase my vcf using SHAPEIT and it requires a genetic map. The first few lines of the example file looks like this: position COMBINED_rate(cM/Mb) Genetic_Map(cM) 61795 0.3516610754 0 63231 0.3500036909 0.0005026053001324 63244 0.3494018702 0.000507147524445 63799 0.3501262382 0.000701467586646 64150 0.3558643956 0.0008263759895016 64934 0.3567249058 0.0011060483156488 65288 0.3633379498 0.001234669949878 66370 10.0482361599 0.0121068614748898 The genetic map constructed...
-
Dear Xuemei, Yes, 00 is missing data. sexAveraged=1 only puts average map position to the output maps. It does not change the algorithms in any way. The same effect can be obtained by calculate the average map position from the output of LM3 with sexAveraged=0. Cheers, Pasi
-
Xuemei Tang posted a comment on discussion General Discussion
Hi Pasi, Thanks for your quick response! simpleConvert.awk works perfectly. I would like to check “00” means missing data? When sexAveraged=0, LM3 use all markers to generate male/female position? e.g. for male position, use only male informative or male only + both? I know users do not need to set these parameters, but I would like to understand how Lep-MAP3 works theoretically. Best, Xuemei
-
Dear Xuemei, The data is in the input file (p_fil.call)? If you want to call genotypes from it, there is the simpleConvert.awk script to make the calls. LM3 uses all markers as default. You can use informativeMask parameter to use only some subsets of them (e.g. only male informative informativeMask=1, male only + both=13, female only=2, etc.). Cheers, Pasi
-
Hi Pasi, I tried running SeparateChromosomes2 with the reference genome chromosome information, but it didn’t work (it shows 12 linkage groups instead of 9). Please let me know what I did wrong. Thanks for your help! Xuemei java -cp /programs/Lep-MAP3/bin SeparateChromosomes2 data=p_fil.call usePhysical=1 0.01 map=physical.txt > map_physical.txt 2>error_phy.txt awk '(NR>=7)' p_fil.call | cut -f 1,2 >snps.txt paste snps.txt map_physical.txt | cut -f 1-3 > map_physical_lgs.txt error_phy.txt Loading...
-
Hi Pasi, Thanks a lot for your answer! I have two more questions: 1. Is there any way to export genotype data without imputation and correction? 2. For the female and male genetic maps, LM3 only uses the markers that are informative in the female/male parent, or use markers informative in female/male parent and both parents? eg, for female map, only use female informative markers or use female informative and both parents informative markers? Best, Xuemei
-
Hi Pasi, Thanks a lot for your answer! I have two more questions: 1. Is there any way to export genotype data without imputation and correction? 2. For the female and male genetic maps, it only uses the markers that are informative in the female/male parent, or use markers informative in female/male parent and both parents? eg, for female map, only use female informative markers or use female informative and both parents informative markers? Best, Xuemei
-
Dear Xuemei, The usePhysical will use the contig/pos information from your markers as well as the calculated LOD scores between markers. It will decrease calculated LOD value between markers from different contigs (chromosomes) or optionally between markers too far away in the same contig. This is to obtain linkage groups with the help of LOD scores and the physical location of markers. This might be useful if you have a very few individuals in the mapping cross but a good genome. Cheers, Pasi
-
Dear Xuemei, The usePhysical will use the contig/pos information from your markers as well as the calculated LOD scores between markers. It will increase calculated LOD value between markers from different contigs (chromosomes) or optionally between markers too far away in the same contig. This is to obtain linkage groups with the help of LOD scores and the physical location of markers. This might be useful if you have a very few individuals in the mapping cross but a good genome. Cheers, Pasi
-
Hi Pasi, Sorry to bother you again! I would like to know how to set and use the parameter usePhysical. I tried usePhysical=1 0.0001, but it didn’t change the SeparateChromosomes2 results. How exactly does usePhysical work in SeparateChromosomes2 and OrderMarkers2? usePhysical=1 NUM Use physical positions in the marker separation, [not set, NUM=0.01] penalise markers in different contigs by log10(NUM) usePhysical=NUM1 NUM2 (NUM1>1) Use physical positions in the marker separation, penalise markers...
-
Dear Felix, proximityScale is in OrderMarkers2. Cheers, Pasi
-
Felix posted a comment on discussion General Discussion
Dear Pasi, Where is the parameter proximityScale ? I can ont find it in java -cp Filtering2.
-
Dear Xuemei, I think with lg=0, the SeparateChromosomes2 command does not do anything for the map file. Just use your physical.txt as the map file. Cheers, Pasi
-
Hi Pasi, Thanks a lot for your answer! I set lg=0 and was able to get 9 LGs by using reference genome chromosome information (physical.txt). Best, Xuemei java -cp /programs/Lep-MAP3/bin SeparateChromosomes2 data=p_fil.call map=physical.txt lg=0 > map_physical.txt Loading file No grandparents present in family FLBG_0502_FLBG_0503 Number of individuals = 150 Number of families = 1 File loaded with 705 SNPs Number of individuals = 150 excluding grandparents Number of families = 1 computing pairwise...
-
Dear Xuemei, Ah, now I get it. I did not look too carefully how you run Lep-MAP3. Indeed, if you use SeparateChromosomes2 with map parameter, it will run the marker separation on LG1 (parameter lg, default = 1), so only the chr1 is affected. The usePhysical will use the contig/pos information from your markers as well as the calculated LOD scores between markers. Maybe you want something like this (I don't know easier way to achieve this): #physical.txt has N for chrN, for each line cp physical.txt...
-
Hi Pasi, Thanks for your quick response! I thought SeparateChromosomes2 will use reference genome chromosome information (physical.txt) to decide the number of LGs. Xuemei
-
Dear Xuemei Tang, Thank you for your question. This is a very common question, there are many similar threads in the discussion. These could be helpful. What I understand from your result is that the linkage groups 1-9 correspond to chrs1-9. Then there are linkage groups 10-12 with all markers in chr1. So for some reason these markers split into their own groups. If groups 10-12 are very small, you can ignore them. If you are sure that they are in chr 1, just edit the map so that 10-12 => 7. You...
-
Hi Pasi, I tried running SeparateChromosomes2 with the reference genome chromosome information, but it didn’t work (it shows 12 linkage groups instead of 9). Please let me know what I did wrong. Thanks for your help! Xuemei java -cp /programs/Lep-MAP3/bin SeparateChromosomes2 data=p_fil.call usePhysical=1 0.01 map=physical.txt > map_physical.txt 2>error_phy.txt awk '(NR>=7)' p_fil.call | cut -f 1,2 >snps.txt paste snps.txt map_physical.txt | cut -f 1-3 > map_physical_lgs.txt physical.txt 1 161046...
-
Hi Pasi, I tried running SeparateChromosomes2 with the reference genome chromosome information, but it didn’t work (it shows 12 linkage groups instead of 9). Please let me know what I did wrong. Thanks for your help! Xuemei java -cp /programs/Lep-MAP3/bin SeparateChromosomes2 data=p_fil.call usePhysical=1 0.01 map=physical.txt > map_physical.txt 2>error_phy.txt paste snps.txt map_physical.txt | cut -f 1-3 > map_physical_lgs.txt physical.txt 1 161046 ... 1 240003789 2 9757318 ... 2 233285822 3 6187651...
-
Lep-MAP3 updated /binary+code.zip
-
-
Rastas Pasi M A committed [ac6bab] on Code
ProcessCounts
-
Dear Cheng, Thank you for your question. Lep-MAP3 should be very robust on large number of markers. How long the cM distances are? Typically map lengths are in the range of 50-200cM per LG. However, for some species there can be longer LGs. What I suggest first is to check the pedigree structure. For example calculate IBD (module IBD) values between parents and offspring or between offspring from the same families. Also check for outliers for the number of crossovers per individual, these are output...
-
Dear Esteban, Thank you for your message! Yes, you can probably construct linkage map from this data. However, you have to provide a pedigree with full-sib families (both parents). You just don't provide any data for the parents you do not have. I think the error you get indicated that the pedigree is missing the parents. The 0, 1, 2 genotypes must be formatted to Lep-MAP3 likelihoods, assuming 0 and 2 are homozygotes: 0 => 1 0 0 0 0 0 0 0 0 0 1 => 0 1 0 0 0 0 0 0 0 0 2 => 0 0 0 0 1 0 0 0 0 0 These...
-
Cheng Chen posted a comment on discussion General Discussion
Hi Pasi, I am using LEP-MAP3 to construct a genetic map for a Vitis (232 samples, with 500Mb genome size, 19 chromosomes) interspecific population. The resulting map appears unusually long in genetic distance (cM), and I suspect this may be due to the very large number of SNPs I am using (about 320,000). I noticed that the example vcf file only contains ~ 2000 SNPs. Could you please advise on recommended approaches for filtering SNPs to retain only the most informative markers for accurate mapping?...
-
Esteban posted a comment on discussion General Discussion
Dear Pasi, I have a population of half-sib trees, but the parental genotypes are not available, and I would like to perform a linkage mapping analysis using this software. Is it possible to do this with such data? If so, how should I structure the pedigree file? I have tried, but I get the following error: Error 520 — Error: No parent(s) in a family CHR:0. Also, can I use a genotype panel coded as 0, 1, 2, or should it be formatted differently? Thank you very much for your help. Best, Esteban
-
Zach posted a comment on discussion General Discussion
Hi Pasi, Thanks so much for your response. I haven't tried plotting a Marey map, as the genome I am working with is scaffold-level. I tried OrderMarkers2 without grandparentPhase and with proximityScale=100, and that has made a huge difference. Now only one out of 18 linkage groups has any error-indicating red lines in the LMplot output, whereas before it was most of the linkage groups. I tried adding grandparentPhase while running evaluateOrder on these ordered maps, and the likelihood score dropped,...
-
Dear Zach, Thank you for your question. And sorry for taking some time to answer. Are the other linkage groups working fine? Do you have the genome? Have you tried plotting a Marey map? Can you construct the maps without grandparents (grandparentPhase)? Also changing the scale parameter might help (this is the prior for map length, default is 100cM). Sometimes the long non-recombining regions cause problems (this would be visible from the Marey). Also check the marker density, even density is preferred....
-
Hi Pasi and Lep-Map3 community, Thank you for this excellent software and your commitment to answering questions from users. I am looking for advice on how to resolve some concerning LMplot output (see attached). Any explanations you could provide for what these plots are indicating would be very helpful. I understand the red lines indicate errors, but I am unsure how to interpret what these errors mean exactly and what could be causing them. I am also attaching code I used to generate these linkage...
-
Dear Jessica, Thank you for your question. This is probably due to families not being informative on every marker. And you have very long low recombining region as well here making the problem very apparent. Possible solutions: 1) evaluate the map in the physical order of markers 2) use FitStepFunction in Lep-Anchor (Rec) to fix the maps (can use intervals, see 4) 3) filter markers based on how many informative families there are (Filtering2: familyInformativeLimit) 4) ouput the marker intervals...
-
Jessica Hintzsche posted a comment on discussion General Discussion
I'm seeing a strange pattern when I combine families. When I plot male vs. female map positions for a given LG (like in the example below for LG23), I sometimes see parallel bands in the regions with little to no recombination. This only appears when I combine multiple families - I'm working with 9 families (from 7 males and 3 females). When I look at maps for individual families, I don't see this pattern at all. Any advice to resolving these markers?
-
Angus Davison posted a comment on discussion General Discussion
Thanks!
-
Dear Eric, I think the problem is the map file. PlaceAndOrientContigs does not use phase and/or segregation patterns of the map (" (, GA,--, ), 11001..."). You can give the interval map or just (average) map position for each marker. (flags noChromosome=1 and noIntervals=1 might be needed depending how you format your data). I am not quite sure what your code does, but if it inverts the correct regions I am happy with it. You can probably manually make an agp file fixing the inversions and convert...
-
Dear Eric, I think the problem is the map file. PlaceAndOrientContigs does not use phase and/or segregation patterns of the map (" (, GA,--, ), 11001..."). You can give the interval map or just (average) map position for each marker. (flags noChromosome=1 and noChromosome=1 might be needed depending how you format your data). I am not quite sure what your code does, but if it inverts the correct regions I am happy with it. You can probably manually make an agp file fixing the inversions and convert...
-
Dear Eric, java PlaceAndOrientContigs findContigErrors=1 map=map.txt bed=bed.txt Cheers, Pasi
-
Eric Gonzalez posted a comment on discussion General Discussion
Many thanks Pasi, I’m sorry, I’m not entirely sure how to use the FindContigErrors flag. These are the options I can see: java -cp bin FindContigErrors usage: java FindContigErrors [options] chain=file chain file paf=file load alignment file in paf (minimap2) format proximity=file NUM1 NUM2 NUM3 load proximity data, NUM1=bin size [10000] NUM2=max distance in bins [25], NUM3=scale score [1.0] I think I can generate the chain file using the steps described here: https://sourceforge.net/p/lep-anchor/wiki/Home/#creating-alignment-chains-for-haplotype-removal...
-
Dear Eric, Indeed, FitStepFunction assumes that the physical order is correct. If you have inversions, Lep-Anchor has findContigErrors flag that finds such inversions and errors and the new physical order. Then you can use FitStepFunction again in the corrected order... Cheers, Pasi
-
Thank you very much, Pasi. I’m finally getting back to this project. I used the FitStepFunction, but I noticed that it is not working as I expected. For example, I’m attaching one chromosome that I believe shows a very clear inversion. Before and after applying FitStepFunction, it’s evident that the function just removes the linkage information inside the inversion. For large, clear inversions like this one, I could flip them manually, but for smaller ones it might be harder to detect. So, I’d like...
-
Dear Angus, The second column of the input data must be numeric in order for the usePhysical to work. So the format with contig and pos as the first two columns should work. Cheers, Pasi
-
Hi Pasi I am struggling to include the usePhysical command in making a linkage map. Could you clarify the precise format? I have tried the following format (and various others) but without success. Taken a look but cant find precise instructions in the documentation OZ073235.1_85384 OZ073235.1 85384 OZ073235.1_154247 OZ073235.1 154247 OZ073235.1_192828 OZ073235.1 192828 OZ073235.1_225077 OZ073235.1 225077 OZ073235.1_226497 OZ073235.1 226497 Exemplar command is java -Xmx5g -cp /gpfs01/home/plzad/lep-map3-code/bin...
-
Hi Pasi, Thank you very much for the reply. OK now I understand why my F1 data wasn't used since there are no grandparents for them. However they are also not selfing data. So I guess the only way to include them is to not use any phasing information and let the program try all four possibilities, in a separate map? Noted "that male and female map positions make sense only on non-selfing crosses (F1 and F2).". Since my species is actually hermaphrodite and I do not expect any achiasmatic meiosis,...
-
Hi Pasi, Thank you very much for the reply. OK now I understand why my TS1 data wasn't used since there are no grandparents for them. However they are also not selfing data. So I guess the only way to include them is to not use any phasing information and let the program try all four possibilities, in a separate map? Noted "that male and female map positions make sense only on non-selfing crosses (F1 and F2).". Since my species is actually hermaphrodite and I do not expect any achiasmatic meiosis,...
-
Dear Huan, The grandparentsPhase=1 works both with selfing and non-selfing data (when you have grandparents). It will remove parts of the data without grandparents. If you have such mixed types of families, you probably have to make separate maps for different families. (Lep-MAP3 could made to work with these different phasing types, but currently you cannot input the variable phasing information for different families. ) Also note that male and female map positions make sense only on non-selfing...
-
Hi Pasi, In my current dataset, I have several "families" and yes some of them are selfing data, but there are also crossings between F1s. Please see a demo.ped file attached: F1: Grandparents and several F1s including P1 and P2. F2_1: crossing between P1 and P2 where P1 is male (pollen) and P2 is female. F2_2: still crossing between P1 and P2 where P2 is male (pollen) and P1 is female. Here I added duplicated P1 and P2 in the vcf (P1.dup and P2.dup) so their sex code does not conflict with P1 and...
-
Hi Pasi, In my current dataset, I have several "families" and yes some of them are selfing data, but there are also crossings between F1s. Please see a demo.ped file attached: F1: Grandparents and several F1s including P1 and P2. F2_1: crossing between P1 and P2 where P1 is male (pollen) and P2 is female. F2_2: still crossing between P1 and P2 where P2 is male (pollen) and P1 is female. Here I used P1.dummy and P2.dummy so their sex code does not conflict with P1 and P2 in F2_1. F2_3: selfing of...
-
Hi Pasi, In my current dataset, I have several "families" and yes some of them are selfing data, but there are also crossings between F1s. Please see a demo.ped file attached: F1: Grandparents and several F1s including D1 and D2. F2_1: crossing between D1 and D2 where D1 is male (pollen) and D2 is female. F2_2: still crossing between D1 and D2 where D2 is male (pollen) and D1 is female. Here I used D1.dummy and D2.dummy so their sex code does not conflict with D1 and D2 in F2_1. F2_3: selfing of...
-
Dear Huan, Yes, chromosome parameter refers to the linkage group numbers in your map5_js.txt file. Currently, there is no way to have both grandparent and normally phased markers, so you should use selfingPhase=1 with your data (your data was selfing, right?). Cheers, Pasi
-
Dear Huan, Yes, this is expected and possible behaviour. If the LOD score difference between best and second best group is larger than the lodDifference and higher exceed lodLimit, marker is put into the higher LOD group. Without lodDifference, if both LODs are higher than lodLimit, marker is not put to the map. Cheers, Pasi edited: lodLimit=>LodDifference
-
Dear Huan, Yes, this is expected and possible behaviour. If the LOD score difference between best and second best group is larger than the lodDifference and both exceed lodLimit, marker is put into the higher LOD group. Without lodDifference, if both LODs are higher than lodLimit, marker is not put to the map. Cheers, Pasi edited: lodLimit=>LodDifference
-
Dear Huan, Yes, this is expected and possible behaviour. If the LOD score difference between best and second best group is larger than the lodDifference, marker is put into the higher LOD group. Without lodDifference, if both LODs are higher than lodLimit, marker is not put to the map. Cheers, Pasi edited: lodLimit=>LodDifference
-
Thank you very much Pasi for the clarification. I am really glad that I asked. Now I understand the results better! I am current at the OrderMarkers2 step and here are some of my observations/speculations that I hope to verify with you. Supposedly I ran java -cp bin/ OrderMarkers2 map=map5_js.txt data=data_f.call chromosome=1 >order1.txtthe chromosome=1 option is referring to the assignment in map5_js.txt, not the first column in data_f.call, is it? I have multiple families in my pedigree file; one...
-
Dear Huan, Yes, this is expected and possible behaviour. If the LOD score difference between best and second best group is larger than the lodLimit, marker is put into the higher LOD group. Without lodLimit, if both LODs are higher than lodLimit, marker is not put to the map. Cheers, Pasi
-
Just to report on my experience. I also used sed and introduced some problems. Better stick with the awk one-liner that Pasi suggested.
-
Hi Pasi, sorry more questions. The SeparateChromosomes2 yield very nice results, matching the number of chromosomes I have nicely. I am currently at the JoinSingles2All stage where I am trying to put the singletons back to the chromosomes. I have been playing with lodLimit and lodDifference. From what I understand, the lower they are, the less singletons will remain. The results agrees with this assumption, except that when I tried lodDifference = 0 (the default setting I believe), there were more...
-
Thank you Pasi for the response. I should have read the options more carefully! lg=NUM refine linkage group lg [1 if map is provided] I will explore the samplePairs as well. Thanks for the advice. Best, Huan
-
Dear Huan, With the map parameter, SeparateChromosomes2 computes LOD scores only for one linkage group of the map (default lg=1). So there are much fewer markers and therefore it is faster. You can also yield good speedups with samplePairs parameter. For million markers I typically use samplePairs=0.1 to achieve 10x speedup. (I have used 0.01 as well to get results even faster). When this parameter is set, SC2 will sample pair-wise comparisons with this rate. Cheers, Pasi
-
Hi Pasi, I was following your suggestions in this post "https://sourceforge.net/p/lep-map3/discussion/general/thread/48bc87a0e9/" to split my largest group (group1). I used the map I got using lodLimit=25 and set the lodLimit to be 30: SeparateChromosomes2 numThreads=10 data=p.call lodLimit=30 sizeLimit=100 map=LOD25.map.txt > LOD30.map25.txt then I compared its result to just using lodLimit=30 (without any map): SeparateChromosomes2 numThreads=10 data=p.call lodLimit=30 sizeLimit=100 > LOD30.txt...
-
Dear Huan, Yes, 0s are the singletons. Cheers, Pasi
-
Thank you Pasi. One more question, in the output of SeparateChromosomes2, does 0 means singletons (not assigned to any LGs)? Best, Huan
-
Dear Huan, Seems to work correctly. Happy mapping! Cheers, Pasi
-
Hi Pasi, Thank you very much for your prompt reply! It is really rare nowadays to have such helpful and reachable developers! I didn't know it was OK to have individuals in the pedigree but missing in the vcf. Just to make sure I understand you correctly, I made a dummy vcf with the .ped you suggested and run ParentCall2 and this is what I got: Warning: Different number of grandparents (4 and 2) in family F Found 2 grandparents in family F Number of individuals = 6 Number of families = 1 Warning:...
-
Dear Huan, Thank you for your question. You have to add two dummy parents anyway. Just make a F2 pedigree with "dummy1" and "dummy2" as parents and your selfing population as their offspring. You can add the grandparents for dummy1 and dummy2. I have attached a small example pedigree. Please read the wiki carefully how to put the phasing parameters in OrderMarkers2. Cheers, Pasi
-
Dear Pasi, I have a selfing population where I have the genotypes of their grandparents but unfortunately not the parent. On wiki it says "Data for the single parent is not really needed, but the grandparents (say two individuals from different lines crossed to form the parent) can be used. ". How does it work? Do I put the grandparents as parents in the pedigree? Thank you. Best, Huan
-
Dear Cheng, Thank you for sharing your data. I looked it briefly. As you have only 34 offspring, it is a bit difficult to get the linkage groups separated. Especially when you don't have too many markers informative on both parents. Due to this, you might want to construct male and female maps separately (pseudo testcross). I would try one of these ideas: 1) Use physical information to assign markers into linkage groups: awk 'BEGIN{print "#"}(NR>=8){print $1}' v2.ft.call.out >map.chr Use map.chr...
-
Hi Pasi, Thank you for your reply. I’ve tried running the process again, but the results were similar. I noticed that the OrderMasker output for Chromosome 1 is significantly larger than for the others (98 Mb vs. ~1 kb). I checked the ParentCall results, and the SNPs appear to be normally distributed across different chromosomes — is that correct? It seems like the linkage groups might not have been properly assigned to their respective chromosomes. When I reviewed the SeparateChromosomes2 map file,...
-
Dear Cheng, Thank you for your question. These linkage groups do not have markers informative on the father. Do you get a correct number of linkage groups (19?). It seems there are very few markers informative on both parents, only a few tens out of thousands: cut -f 4 order.chr*.txt |sort|uniq -c|sort -n .... 3 ( AC,CA ) 7 ( CA,AC ) 10 ( AC,AC ) 16 ( CA,CA ) 726 ( --,CA ) 1126 ( CA,-- ) 1969 ( AC,-- ) 2218 ( --,AC ) This probably explains your result. As there are no markers informative on both...
-
Hi Pasi, I am using LEP-MAP3 orderMarkers2 to construct a genetic map. The population I'm working with is derived froman interspecific cross between two species within Vitis. However, I noticed that the marker positions for one parent are completely missing in some chromosome of the output files. I'm not very familiar with genetics, so I would really appreciate your help in understanding where the issue might be. #java OrderMarkers2 map=map5_js.txt data=data_f.call chromosome=14 randSeed=-2790967626678564236...
-
Dear Lorenzo, Thank you for your question. It is the fact that all markers are not informative on all families. This causes uncertainty in the map positions (see my Lep-Anchor paper, "Such uncertainty can occur due to low recombination regions, cross type (e.g. multi-family data) or poor mapping data (genotype) quality."). You can get the marker position intervals (calculateIntervals) that have the uncertainty information. You can then use this with the maps in FitStepFunction (Lep-Rec). Or you can...
-
Dear Eric, It is the noChromosome=1 flag that is needed... (I just updated FitStepFunction so that it will autodetect whether chromosome is present or not in the input...) FitStepFunction is easiest to use with the raw maps from OrderMarkers2 and SNP file as: java FitStepFunction map=map_from_om2.txt snps=snps.txt Cheers, Pasi
-
Dear Eric, It is the noChromosome=1 flag that is needed... FitStepFunction is easiest to use with the raw maps from OrderMarkers2 and SNP file as: java FitStepFunction map=map_from_om2.txt snps=snps.txt Cheers, Pasi
-
Hello Pasi, We are building linkage maps with LepMap3 using 9 families (from 7 males and 3 females). Data is highly accurate (SNP Chip) and we have1,962 offspring in total (smallest family has 99 offspring). When we build maps from single families they look really good (1:1 match between male and female map, no off diagonal markers). But when we join multiple families, we start seeing 'parallel lines' (see attached image for an example on one chromosome). Notably, the map with this issue has the...
-
Many thanks for your answer. I would like to use use the FitStepFunction from Lep-Rec. I am trying to do this: java -cp /LEP-REC/bin FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped But I am getting this error: java FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped java.lang.NumberFormatException: For input string: "0.000" at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67) at java.base/java.lang.Integer.parseInt(Integer.java:668)...
-
Many thanks for your answer. I would like to use use the FitStepFunction from Lep-Rec. I am trying to do this: java -cp /LEP-REC/bin FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped But I am getting this error: java FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped java.lang.NumberFormatException: For input string: "0.000" at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67) at java.base/java.lang.Integer.parseInt(Integer.java:668)...
-
Many thanks for your answer. I would like to use use the FitStepFunction from Lep-Rec. I am trying to do this: java -cp /LEP-REC/bin FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped But I am getting this error: java FitStepFunction map=order_LG1_informativeMask1_reevaluated.mapped java.lang.NumberFormatException: For input string: "0.000" at java.base/java.lang.NumberFormatException.forInputString(NumberFormatException.java:67) at java.base/java.lang.Integer.parseInt(Integer.java:668)...
-
Yutang Chen posted a comment on discussion General Discussion
Thank you so much for your great explanation, Pasi. I will try my best to digest what you wrote. Best wishes, Yutang
-
Dear Yutang, ...the male and female map would correspond to the map of P1 and P2, respectively, right? Yes For the second part, the first map position is P1 map and second P2 map. However, the markers are in the same order in both of these. If this is not a problem (e.g. due many inversions) I don't think there is a problem calling then map of P1 and P2. And lastly, Lep-MAP3 separates parental haplotypes based (mostly) on the markers where only one parent is informative. Then the markers where both...
1 >
