DReichLab / EIG

Eigen tools by Nick Patterson and Alkes Price lab
Other
180 stars 60 forks source link

please explain #23

Open mike2vandy opened 6 years ago

mike2vandy commented 6 years ago

~/betterSoftware/EIG/bin/smartpca.perl \ -i samples.eigen.geno \ -a samples.eigen.snp -b samples.eigen.ind \ -o test.pca \ -p test \ -e test.eigan.eval \ -l test.log

smartpca -p test.pca.par >test.log fatalx: bad chrom: LT635612.1 sh: line 1: 4589 Aborted (core dumped) smartpca -p test.pca.par > test.log ploteig -i test.pca.evec -c 1:2 -p -x -y -o test.xtxt evec2pca.perl 10 test.pca.evec samples.eigen.ind test.pca

bumblenick commented 6 years ago

Alkes Price's lab maintains the perl script but I'll bet that samples.eigen.snp is not in reichlab format. Chromosome values are {1-22, X, Y} and are in the second column. snpnames (like LT635612.1 ??) are in the first column. See examples.

Nick

On Wed, Feb 7, 2018 at 4:10 PM, mike2vandy notifications@github.com wrote:

~/betterSoftware/EIG/bin/smartpca.perl -i samples.eigen.geno -a samples.eigen.snp -b samples.eigen.ind -o test.pca -p test -e test.eigan.eval -l test.log

smartpca -p test.pca.par >test.log fatalx: bad chrom: LT635612.1 sh: line 1: 4589 Aborted (core dumped) smartpca -p test.pca.par > test.log ploteig -i test.pca.evec -c 1:2 -p -x -y -o test.xtxt evec2pca.perl 10 test.pca.evec samples.eigen.ind test.pca

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/DReichLab/EIG/issues/23, or mute the thread https://github.com/notifications/unsubscribe-auth/AQn_h0y0cCchCut3CX0dRxP999s9cl38ks5tShEugaJpZM4R9ZVM .

mike2vandy commented 6 years ago

So if my .snp file looks like:

snp#0 LT635612.1 6e-06 6 G T snp#1 LT635612.1 7e-05 70 A C snp#2 LT635612.1 9.2e-05 92 T A snp#3 LT635612.1 9.3e-05 93 A T snp#4 LT635612.1 0.000121 121 C T snp#5 LT635612.1 0.000124 124 T C snp#6 LT635612.1 0.000125 125 T C snp#7 LT635612.1 0.000131 131 A T snp#8 LT635612.1 0.000148 148 G A snp#9 LT635612.1 0.000159 159 A T

Can eigensoft not recognize chromosome/contigs names != {1-22, X, Y}? Can that be fixed?

bumblenick commented 6 years ago

Nope: Your C-names won't parse. You will have to find a way to change them.

N

On Fri, Feb 9, 2018 at 1:15 PM, mike2vandy notifications@github.com wrote:

So if my .snp file looks like:

snp#0 LT635612.1 6e-06 6 G T snp#1 LT635612.1 7e-05 70 A C snp#2 LT635612.1 9.2e-05 92 T A snp#3 LT635612.1 9.3e-05 93 A T snp#4 LT635612.1 0.000121 121 C T snp#5 LT635612.1 0.000124 124 T C snp#6 LT635612.1 0.000125 125 T C snp#7 LT635612.1 0.000131 131 A T snp#8 LT635612.1 0.000148 148 G A snp#9 LT635612.1 0.000159 159 A T

Can eigensoft not recognize chromosome/contigs names != {1-22, X, Y}? Can that be fixed?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/DReichLab/EIG/issues/23#issuecomment-364513950, or mute the thread https://github.com/notifications/unsubscribe-auth/AQn_h1FrEmHaDTfz_vhv0nAcdS42jy90ks5tTIs4gaJpZM4R9ZVM .

mike2vandy commented 6 years ago

Imagine if BWA were built this way...

bumblenick commented 6 years ago

Feel free to rewrite the software!

Nick

On Fri, Feb 9, 2018 at 1:35 PM, mike2vandy notifications@github.com wrote:

Imagine if BWA were built this way...

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/DReichLab/EIG/issues/23#issuecomment-364519330, or mute the thread https://github.com/notifications/unsubscribe-auth/AQn_h68cSwQ3La6nmEvh3OH3mXU3ICm0ks5tTI_4gaJpZM4R9ZVM .