Perseus annotation help

[Sílvia Batista]

Hi, 

I am quite new using Perseus. I have downloaded annotation fasta file, added to perseus, conf, annotations and when on Perseus I click add annotation it appears there, I selected KEGG, corum, etc and added. 

All looked ok but categorical columns are completely empty with no annotations in it when new matrix is generated. Any idea what might have I done wrong??

I even tested a dataset that we used on MaxQuant Summer school this year to be sure was not my data but still the same problem......

Thanks in advance

[Und Valle So Yeah]

Hey, I got the same problem. Do you know what the problem was?

Thanks.

[Kate Samardzic]

I have also had this problem. I have downloaded the mainAnnot.homo_sapiens.txt.gz file and put in conf/annotations but still nothing when the new matrix is generated? 

Has anyone resolve this?

Thanks,

Kate

[Animesh Sharma]

Hi Kate,

Which version of Perseus are you using (mine is 1.6.1.3). If i have to guess the issue, probably it is because of issues with the setting up the annotation itself? The example at http://coxdocs.org/doku.php?id=perseus:user:activities:MatrixProcessing:Annotcolumns:AddAnnotationToMatrix is for Chromosomal annotations using Ensemble ID. 

First make sure that "Protein IDs" column has Human Uniprot IDs like "S4R3N1;Q9Y3A3-2;Q9Y3A3-3;Q9Y3A3;B4DM50" (individual one separated by ";") for GO assuming Human Uniprot ID and then change defaults to:

All the best!

Regards,

Ani


--------------------------"The Answer Lies In The Genome"--------------------------

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

-rms

--
You received this message because you are subscribed to the Google Groups "Andromeda Help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to andromeda-list+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[bal...@colorado.edu]

A couple of things that you can check:

1. is the annotation file properly unzipped?

2. if it is unzipped properly; you should be able to open the annotation file with notebook and it should look a bit like this:

Entry	Is_kinase	Entry name	Gene names
#!{Type}	Categorical	Text	Text
Q59H18	yes	TNI3K_HUMAN	TNNI3K CARK

...

(I make a lot of custom annotation files: that #!{Type} is the key to making work!)  

Make sure that in Perseus's annotation ENSG column drop down menu that you select YOUR column that matches the #!{Type}. In my example, I used the UniprotID as the ID type. Other options are protein name or gene name, but I like Uniprot the best; The Perseus standard annotation files like the uniprotID best too!  The default ENSG column selection is the top text type data; it doesn't recognize which column is consistent with a uniprotID.  This is usually the problem.

3. Next step, all the columns in the annotation file after the #!{Type} column are annotations that COULD be used; and they are in the left Annotations Box.  You have to select the ones you want added! Select and  [>] key to move the annotations you want to the right box. Only the annotations in the right box will be added.

4. Don't be greedy!  If you select 2-3 annotations your annotations will be done in a few minutes; if you select 20-30 it could be hours and something is likely to crash before it is done! Add 2-3 then add 2-3 more...

5. If you have Categorical annotations then if make a volcano plot or scatter plot you can go to the categorical and select a specific category to highlight on your plot. With my example annotation file - I can highlight all the kinases!

If you test all these things and it doesn't work try making a simple annotation file to test using your uniprot IDs (symbols ie ;) you can download a short ID to gene name list from Uniprot's Ritrieve/ID mapping.

Then just add row under the column names and add the 

#!{Type}

Categorical

Text

descriptors (sames data types that Perseus uses for the data upload).

My custom annotation files have worked on the last 3-4 versions of Perseus; 1.6.1.3 at the time I posted this.

Good Luck

[MIN MA]

Hi everyone:

I have the same issue for any version on my desktop. I have checked the annotation files. The looks like the same as you mentioned.

But, I cannot get any results after adding annotation.

Can anyone have other solutions to solve this issue?

Thank you so much in advance for your help.

Thanks,

Min

[Colin Hammond]

I think this is to do with the formatting of the uniprot database you use during the preceding analysis, e.g. in maxquant. If you don't get a column containing uniprot accession codes in your proteinGroups file (e.g. Q71U36 without isoform information, no dashes and numbers after) then the catagorical annotation doesn't work. At least for me this didn't work. You can extract the uniprot accession codes for the canonical isoform from Majority protein IDs into a new column. Then use this column as the source uniprot column in the catagorical annotation. Then select the annotations you want to add. It should work then. 

Hope this helps,

Colin

--

You received this message because you are subscribed to the Google Groups "Andromeda Help" group.

To unsubscribe from this group and stop receiving emails from it, send an email to andromeda-lis...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/andromeda-list/6feb449b-d99c-40fe-a937-5ed50b4df510%40googlegroups.com.

[Michael Regan]

Hi everyone, 

I'm very new to Perseus and require the addition of annotations to move to the next stage of a research project. Unfortunately, the annotations seem to be crashing Perseus even if taken one at a time. There are three that can be added to my matrix one by one, but the remaining seven of those I require constantly return this error message after about five minutes or so:

Has anyone had any luck resolving an issue like this?

It would be greatly appreciated!

Thanks everyone :)