genotyping code of qtl_cartographer data for BC or RILs
56 views
Skip to first unread message
andbelo
Dec 11, 2008, 12:06:39 PM12/11/08
to R/qtl discussion
Dear Karl,
I noticed that when importing qtlcart.cross files from QTLcartographer
and changing the cross type, R/qtl recognizes the heterozygous
genotypes as BB. The two genotype classes will be AA and H, instead of
the expected AA and BB.
Although there is a warning about genotype problems I wonder if it
could be worth to change the way R/qtl recognizes the genotypes.
QTLcartographer uses 1 for A, 0 for H, and 2 for B. When the
population is set to F2 in R/qtl it recognizes the three classes fine,
but as soon as you change it to 'bc' or 'riself', it only gets the
first two: 1 (A) and 0 (H), leaving BB behind. One would expect it
picked the A and B genotypes. For example:
R> d = read.cross('qtlcart', file = 'qtlcart.cro', mapfile =
'qtlcart.map')
--Read the following data:
Type of cross: f2
Number of individuals: 72
Number of markers: 522
Number of phenotypes: 4
--Cross type: f2
R> summary(d)
F2 intercross
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:49.2 AB:2.9 BB:47.9 not BB:0 not AA:0
R> class(d)[1] = 'bc'
R> summary(d)
Backcross
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:94.5 AB:5.5
Warning message:
In summary.cross(d) : Invalid genotypes.
Observed genotypes: 1 2 3
R> class(d)[1] = 'riself'
R> summary(d)
RI strains via selfing
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:94.5 BB:5.5
Warning message:
In summary.cross(d) : Invalid genotypes.
Observed genotypes: 1 2 3
I hope this might be a helpful comment to help out in improving R/qtl.
Thank you.
I noticed that when importing qtlcart.cross files from QTLcartographer
and changing the cross type, R/qtl recognizes the heterozygous
genotypes as BB. The two genotype classes will be AA and H, instead of
the expected AA and BB.
Although there is a warning about genotype problems I wonder if it
could be worth to change the way R/qtl recognizes the genotypes.
QTLcartographer uses 1 for A, 0 for H, and 2 for B. When the
population is set to F2 in R/qtl it recognizes the three classes fine,
but as soon as you change it to 'bc' or 'riself', it only gets the
first two: 1 (A) and 0 (H), leaving BB behind. One would expect it
picked the A and B genotypes. For example:
R> d = read.cross('qtlcart', file = 'qtlcart.cro', mapfile =
'qtlcart.map')
--Read the following data:
Type of cross: f2
Number of individuals: 72
Number of markers: 522
Number of phenotypes: 4
--Cross type: f2
R> summary(d)
F2 intercross
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:49.2 AB:2.9 BB:47.9 not BB:0 not AA:0
R> class(d)[1] = 'bc'
R> summary(d)
Backcross
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:94.5 AB:5.5
Warning message:
In summary.cross(d) : Invalid genotypes.
Observed genotypes: 1 2 3
R> class(d)[1] = 'riself'
R> summary(d)
RI strains via selfing
No. individuals: 72
No. phenotypes: 4
Percent phenotyped: 100 100 98.6 100
No. chromosomes: 10
Autosomes: 1 2 3 4 5 6 7 8 9 10
Total markers: 522
No. markers: 87 48 38 41 76 49 46 66 39 32
Percent genotyped: 96.6
Genotypes (%): AA:94.5 BB:5.5
Warning message:
In summary.cross(d) : Invalid genotypes.
Observed genotypes: 1 2 3
I hope this might be a helpful comment to help out in improving R/qtl.
Thank you.
andbelo
Dec 11, 2008, 1:15:35 PM12/11/08
to R/qtl discussion
More on genotyping code of qtl_cartographer data for BC or RILs:
I found out that it depends on the previous cross type present in the
QTLcartographer file. For example, if the original cross type in
QTLcartographer was RI1, then R/qtl will correctly select the A and B
genotypes. So, it seems that what I described before will only happen
when the type of cross is changed after importing.
I found out that it depends on the previous cross type present in the
QTLcartographer file. For example, if the original cross type in
QTLcartographer was RI1, then R/qtl will correctly select the A and B
genotypes. So, it seems that what I described before will only happen
when the type of cross is changed after importing.
Reply all
Reply to author
Forward
0 new messages