I have some questions regarding consensus map. I wish you could share your brilliant input here with us.
First, some background of what we have done:
We did reciprocal crossing between two heterozygous parents. The parents are allogamous forage grasses, and let's call them P1 and P2.
We collected seeds from both parents, resulting in ~300 F1 individuals.
We sequenced these F1 plants together with the parental plants using genotyping-by-sequencing (GBS), and we just followed your map construction pipeline (with mostly defaut parameters, except the fine-tuned value of lodLimit and theta in SeparateChromosomes2) and generated linkage maps.
Our grasses don't have achiasmatic chromosomes.
Now questions:
The seeds were collected from heads of both parents, in this case, in the pedigree file, how should we assign parents to the F1 individuals? I just assigned P1 as father and P2 as mother to all F1 individuals. Does this matter for our grasses? I attached an example of our pedigree file.
I used all informative markers to build the map. This means the resulting male and female maps included 1) markers that are only informative in each parent and also 2) markers that are informative in both parents. Would you call the male and female maps consensus maps? I considered them as consensus maps because they contain markers that are heterozygous in both parents.
Although the male and female maps had the same markers, and I considered them as consensus maps, I found that the position of markers was different between the two maps (the order of the markers is the same between the two maps). I am wondering what have affected the calculation of the map position between male and female maps? Can I still call them consensus maps or shall I just call them male and female maps?
I found here https://sourceforge.net/p/lep-map3/discussion/general/thread/cee88a47d2/?limit=25#8773/a4fb that sexAveraged=1 could be used to generate the real consensus map between male and female, meaning that the resulting male and female maps will be identical with the same marker and the same position. Could you please tell us a little bit more about what this sexAveraged=1 does?
I am very sorry for asking so many questions. Please don't get stressed by my questions, and I would appreciate your answer very much if you have chance to read my questions.
Look forward to your reply and thank you very much.
Yutang
This is very interesting question. Hope I understood correctly the question.
I guess in this case you can construct parental maps and sex-specific maps.
For the map construction, this does not matter. P1 and P2 are the parents of each individual.
Could you call these parental maps? map of P1 and P2.
For male and female maps you should know which parent is mother and which father for each individual. If you know from which parent each seeds was collected, it would suffice? (I have constructed maps for a hermaphrodite fish, there the mitochondria was used to determine which parent was father and which mother for each individual). You can "simply" calculate the male and female map positions from your maps if you know the parental-sex status of each individual. This can be done by swapping the paternal and maternal crossovers accordingly, there is no need to redo maps.
sexAveraged=1 only averages the map positions. You can do this for your map simply by computing the average of male and female position.
Cheers,
Pasi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Great, Pasi. Thank you so much for your very helpful reply.
Maybe I should have separated the last three questions from my first question to make my last three questions not related to the first one.
For the first question, I totally agree with you. For the 4th question, thanks for elaborating on sexAveraged=1. Now I got it.
Now for question 2 and 3, actually it is not about the sex status of the parents in the pedigree file (the parental-sex status of the individuals was assigned to 0), it is more about how we should call/understand the male and female map in our grass context:
For our grasses, let's just assume we have done everything properly following your pipeline, and now we have got the order.txt file containing the three columns: marker_number, male_position and female_position. So if I understand correctly, the male_position and female_position represent the male and female map, respectively. If, in the pedigree file, we have assigned P1 and P2 as male and female, respectively, then the male and female map would correspond to the map of P1 and P2, respectively, right?
Regarding the P1 and P2 map, I got confused because normally we would consider a P1 map as P1-specific, meaning that P1 map only contains P1-informative markers. Same to the P2 map. However, here, if we have all the informative markers for each map (including markers are only heterozygous in one parent and makers are heterozygous in both parents), I am not sure if I should still call them P1 or P2 map. Would you call these non-specific maps as consensus maps as they contain all informative markers? Or how would you call these maps?
If P1 and P2 have all the same informative markers, then I got confused by why the marker position is different between the two maps. Any ideas to explain this? Is this because the parent-specific markers (markers heterozygous just in one parent) makes a difference during computing? This is a difficult part for us because we don't how the algorithm works.
Thank you very much again Pasi for always being nice and patient to all the users of Lep-Map3. If you find some time to write to me back, I would appreciate it very very much.
Best wishes,
Yutang
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
...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 parents are informative can be separated into haplotypes (and recombinations) of parental maps.
If you don't have single parent informative markers (or multi-allelic markers), then you cannot separate parental haplotypes. Then there can be position differences in Lep-MAP3 just due to randomly assigning the haplotypes into two. Taking the average map position in this case would be best. This is the case for example when analysing selfing crosses.
Cheers,
Pasi
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear Pasi,
Long time no see, and I hope you are well.
I have some questions regarding consensus map. I wish you could share your brilliant input here with us.
First, some background of what we have done:
Now questions:
I am very sorry for asking so many questions. Please don't get stressed by my questions, and I would appreciate your answer very much if you have chance to read my questions.
Look forward to your reply and thank you very much.
Yutang
Dear Yutang,
Nice to hear from you.
This is very interesting question. Hope I understood correctly the question.
I guess in this case you can construct parental maps and sex-specific maps.
Cheers,
Pasi
Great, Pasi. Thank you so much for your very helpful reply.
Maybe I should have separated the last three questions from my first question to make my last three questions not related to the first one.
For the first question, I totally agree with you. For the 4th question, thanks for elaborating on sexAveraged=1. Now I got it.
Now for question 2 and 3, actually it is not about the sex status of the parents in the pedigree file (the parental-sex status of the individuals was assigned to 0), it is more about how we should call/understand the male and female map in our grass context:
For our grasses, let's just assume we have done everything properly following your pipeline, and now we have got the order.txt file containing the three columns: marker_number, male_position and female_position. So if I understand correctly, the male_position and female_position represent the male and female map, respectively. If, in the pedigree file, we have assigned P1 and P2 as male and female, respectively, then the male and female map would correspond to the map of P1 and P2, respectively, right?
Regarding the P1 and P2 map, I got confused because normally we would consider a P1 map as P1-specific, meaning that P1 map only contains P1-informative markers. Same to the P2 map. However, here, if we have all the informative markers for each map (including markers are only heterozygous in one parent and makers are heterozygous in both parents), I am not sure if I should still call them P1 or P2 map. Would you call these non-specific maps as consensus maps as they contain all informative markers? Or how would you call these maps?
If P1 and P2 have all the same informative markers, then I got confused by why the marker position is different between the two maps. Any ideas to explain this? Is this because the parent-specific markers (markers heterozygous just in one parent) makes a difference during computing? This is a difficult part for us because we don't how the algorithm works.
Thank you very much again Pasi for always being nice and patient to all the users of Lep-Map3. If you find some time to write to me back, I would appreciate it very very much.
Best wishes,
Yutang
Dear Yutang,
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 parents are informative can be separated into haplotypes (and recombinations) of parental maps.
If you don't have single parent informative markers (or multi-allelic markers), then you cannot separate parental haplotypes. Then there can be position differences in Lep-MAP3 just due to randomly assigning the haplotypes into two. Taking the average map position in this case would be best. This is the case for example when analysing selfing crosses.
Cheers,
Pasi
Thank you so much for your great explanation, Pasi. I will try my best to digest what you wrote.
Best wishes,
Yutang