Following Tassel 5 user manual: 10 tutorial-10.2 PCA to run TASSEL 5.2 GUI

1,385 views
Skip to first unread message

huana R

unread,
Nov 26, 2014, 11:42:22 PM11/26/14
to tas...@googlegroups.com
Hello,

I am following the tutorial step of PCA analysis according to the latest TASSEL 5.

1) The selectchromosome function doesn't work properly. I need choose chromosome one by one and end up with 7 chromosome  dataset, which I need combine them to one to do the later analysis. I don't know how. I checked each dataset, the number of site all are 64. This shouldn't be right. As I know I have nearly 1000 SNPs in each chromosome.  

2) I used the chr_1 dataset continue to do numericalization. I tried numerical genotype function first and then tried transform. A window come out with the dataset selected is wrong type. As other people mentioned, I couldn't get my numerical genotype dataset. 

3) Then I went to impute-impute by FILLIN. It is quiet confusing to read the manual. Imputing may service different purpose. But I just want to impute the missing genotypes for further study. I don't want to do haplotypes study. Anyway, after running the imputation, the results didn't show on the dataset. But after that, I can run PCA. The only pity is I run PCA only with 64 SNPs from chromosome 1 instead of 5261 SNPs from the whole genome. 

I am going to try your old version which I could get the numerical genotypes. It's very useful for further study. The only problem I think is to fix the result of  filter-sites, which always come back with 64 site numbers, cutting most SNPs out.  

I am a GUI user. 

Thank you,

Huana


Peter Bradbury

unread,
Nov 27, 2014, 1:36:13 PM11/27/14
to tas...@googlegroups.com
Version 5.2 changes the way numerical genotypes and PCA are handled. "Select chromosome" works for me on the tutorial data set. I cannot reproduce your problem with that. If you want to create separate data sets for each chromosome it is easier to use Data/Separate, which does that for all chromosomes at once. Numerical genotypes are stored as ReferenceProbability. You can see those by selecting ReferenceProbability in the dropdown box on the Genotype viewer. However, if you just want to run PCA, you can now do that on genotype data directly in 5.2 (the numerical conversion and imputation is done automatically as part of PCA). Just choose the genotype and use Analysis/PCA. I will let someone else comment about FILLIN.

Peter

huana R

unread,
Feb 7, 2015, 7:47:21 PM2/7/15
to tas...@googlegroups.com
Hello, Peter

Thank you for your message. It's very helpful.

I am a beginner of TASSEL and I have few other questions to ask you. My dataset is whole genome SNP diploid dataset. I tried to study my population structure by using Cladogram_NJ clustering methods and PCA analysis. After that I am going to use STRUCTURE software to confirm my cluster results. So:

1) What is the difference between Kinship with STRUCTURE software? 
2) In the Cladogram analysis, you use a modified Euclidean distance model to to calculate distance matrix. What is the difference between this distance matrix with IBS DistanceMatrix analysis?  Can I use IBS DistanceMatrix to do cluster analysis? If yes, How?


Regards,
Huana






--
You received this message because you are subscribed to a topic in the Google Groups "TASSEL - Trait Analysis by Association, Evolution and Linkage" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/tassel/Q7uqxs3iZvk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to tassel+un...@googlegroups.com.
To post to this group, send email to tas...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tassel/efa02ed1-0f69-479b-b8e5-2511bc59d883%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

huana R

unread,
Feb 7, 2015, 11:34:16 PM2/7/15
to tas...@googlegroups.com
Hi, One more question: 

By selecting ReferenceProbability in the dropdown box on the Genotype viewer, I see the numerical genotypes, which is transformed from .hapmap file.  If I want download the numerical genotypes, what should I do?  The earlier version of TASSEL allowed to save the transformed numerical genotypes, but the latest one didn't.

Thank you. 

Terry Casstevens

unread,
Mar 3, 2015, 3:49:15 PM3/3/15
to Tassel User Group
The ability to export numerical genotypes will be available in the next release.

Best,

Terry
> You received this message because you are subscribed to the Google Groups
> "TASSEL - Trait Analysis by Association, Evolution and Linkage" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to tassel+un...@googlegroups.com.
> To post to this group, send email to tas...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/tassel/CAGkw7guDOEpFs4oiJK%2Bxe5EB9sHwO1vuZk7pLRtfDNn%3DO2X7Ug%40mail.gmail.com.
Reply all
Reply to author
Forward
0 new messages