Manhattan plot issues and other questions

fer...@gmail.com

unread,

Jul 28, 2015, 11:36:55 AM7/28/15

to TASSEL - Trait Analysis by Association, Evolution and Linkage

Hello,

I am still new to Tassel and I can't find answers to these questions:

1. Is there a limit in chr number? I use dog data with 39 chrs and it seems something is wrong. The manhattan plots I get are all weird, like the one in the picture I attached.

I also have chr 0 (zero) for the SNPs for which there isn't enough information in dogs. Could any of these things be interfering in the manhattan plot? Does that mean the GLM and MLM are not reliable?

2. I am trying to use covariates, my data is paired for breed and sex, so I assigned a number for each pair to use as covariate. And I have a score for behaviour phenotype as data in the phenotype file. Is there any way of knowing if the covariate is being correctly integrated in the analysis?

Many thanks,

Fernanda Fadel

manhattanplotfernanda.png

Terry Casstevens

unread,

Jul 28, 2015, 11:44:44 AM7/28/15

to Tassel User Group

Are some of your positions recorded as negative numbers?

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/968af4dd-ac62-4a3e-8672-85566827d596%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Fernanda Fadel

unread,

Jul 28, 2015, 12:29:41 PM7/28/15

to tas...@googlegroups.com

No. The chr are 0 to 39.

--
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/gfxqIDppOao/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/CACHsrTtdKTgabkseYKincUA0XqM7z7wPZrbxp5Z%2B57gFfEkuJg%40mail.gmail.com.

Edward S. Buckler

unread,

Jul 28, 2015, 12:50:38 PM7/28/15

to tas...@googlegroups.com

This look like a integer rollover situation? The whole genome probably adds up to more than what an integer can hold 2^31.

This is a likely a bug, and the x-axis index needs to be converted to a long

Cheers-

Ed

--
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/968af4dd-ac62-4a3e-8672-85566827d596%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<manhattanplotfernanda.png>

Peter Bradbury

unread,

Jul 28, 2015, 12:52:16 PM7/28/15

to TASSEL - Trait Analysis by Association, Evolution and Linkage, fer...@gmail.com, fer...@gmail.com

Breed and sex are categorical data. In Tassel, independent, categorical variables are called "factors". Covariates are continuously variable numbers. Tassel assumes that you think that a covariate and the response variable are correlated. From your description, it sounds like the number you assign to each breed-sex combination should be called a factor.

Peter

Terry Casstevens

unread,

Jul 29, 2015, 11:57:55 AM7/29/15

to Tassel User Group

The Manhattan Plot will be fixed in the next build tomorrow. Let me know if you still see problems.

Best,

Terry

To view this discussion on the web visit https://groups.google.com/d/msgid/tassel/71E515EC-EA53-4286-AB5A-92F27CC9B2A0%40cornell.edu.

Sandip Kale

unread,

Jan 11, 2016, 10:30:20 PM1/11/16

to TASSEL - Trait Analysis by Association, Evolution and Linkage

Hi Terry,

I am using tassel 5 and still facing the same issue. Kindly find attached file. Please let me know, once the issue gets resolved.

Sandip

On Wednesday, July 29, 2015 at 9:27:55 PM UTC+5:30, Terry wrote:

The Manhattan Plot will be fixed in the next build tomorrow. Let me know if you still see problems.

Best,

Terry

On Tue, Jul 28, 2015 at 12:50 PM, Edward S. Buckler <es...@cornell.edu> wrote:

This look like a integer rollover situation? The whole genome probably adds up to more than what an integer can hold 2^31.

This is a likely a bug, and the x-axis index needs to be converted to a long

Cheers-

Ed

On Jul 28, 2015, at 11:36 AM, fer...@gmail.com wrote:

Hello,

I am still new to Tassel and I can't find answers to these questions:

1. Is there a limit in chr number? I use dog data with 39 chrs and it seems something is wrong. The manhattan plots I get are all weird, like the one in the picture I attached.

I also have chr 0 (zero) for the SNPs for which there isn't enough information in dogs. Could any of these things be interfering in the manhattan plot? Does that mean the GLM and MLM are not reliable?

2. I am trying to use covariates, my data is paired for breed and sex, so I assigned a number for each pair to use as covariate. And I have a score for behaviour phenotype as data in the phenotype file. Is there any way of knowing if the covariate is being correctly integrated in the analysis?

Many thanks,

Fernanda Fadel

--
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 ta...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/tassel/968af4dd-ac62-4a3e-8672-85566827d596%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<manhattanplotfernanda.png>

test.png

Terry Casstevens

unread,

Jan 11, 2016, 10:51:16 PM1/11/16

to Tassel User Group

I just checked, and this is working for me. Please see this document
how to report issue...

https://bitbucket.org/tasseladmin/tassel-5-source/wiki/docs/ReportingTassel5Issues.pdf

> https://groups.google.com/d/msgid/tassel/d8045e5e-ca38-4747-9117-852ce9408aed%40googlegroups.com.

xiang lin

unread,

Mar 31, 2016, 7:56:04 PM3/31/16

to TASSEL - Trait Analysis by Association, Evolution and Linkage

Hi Terry,

I have a problem with Manhattan plot. When I plot results from MLM to Manhattan, this is what happened below.

Thank you

Shawn

Manhattan.png

Terry Casstevens

unread,

Mar 31, 2016, 10:45:53 PM3/31/16

to Tassel User Group

This looks like the key for chromosomes. Do you have that many chromosomes?

> https://groups.google.com/d/msgid/tassel/09be417d-d119-4312-a2e6-143a6e9c5942%40googlegroups.com.

xiang lin

unread,

Apr 1, 2016, 11:36:49 AM4/1/16

to TASSEL - Trait Analysis by Association, Evolution and Linkage

Hi Terry,

There are 29 chromosomes. Is there a way that I could change the setup in Tassel for Manhattan Plot. I would like to have chromosomes on x-axis and get rid of all the key (locus information) because that blocks whole manhattan plot.

Terry Casstevens

unread,

Apr 1, 2016, 11:41:02 AM4/1/16

to Tassel User Group

I don't have enough information to answer your question.

https://bitbucket.org/tasseladmin/tassel-5-source/wiki/docs/ReportingTassel5Issues.pdf

Best,

Terry

> https://groups.google.com/d/msgid/tassel/01d00b8a-bfb4-429d-83bf-2d856cbe5ce2%40googlegroups.com.

xiang lin

unread,

Apr 1, 2016, 12:19:39 PM4/1/16

to TASSEL - Trait Analysis by Association, Evolution and Linkage

Hi Terry,

Thank you very much for your quick reply. I tried Manhattan Plot again and changed the P-value on MLM analysis. Attached is the Manhattan Plot. I'm still trying to get rid of the locus information on the plot. Thank you.

Z`8JKP~F%{A963]RK(5YO_O.jpg

Terry Casstevens

unread,

Apr 1, 2016, 12:38:07 PM4/1/16

to Tassel User Group

Can you export and send me the text output that you are using to
create the Mannhattan plot?

> https://groups.google.com/d/msgid/tassel/8e96ba31-faa6-404e-b49e-785257f003d6%40googlegroups.com.

Reply all

Reply to author

Forward