DAVID unable to determine the biological process/ function of gene list

Question

[GreenDiamond]

I have a list of DEGs from DESeq from A.digitifera coral. I put them into DAVID software as a list, selected Identifier as "OFFICIAL_GENE_SYMBOL", chose Gene List, and then submitted the list. Then, I chose "Option 1: Convert the gene list to DAVID". None of the genes were in the DAVID database, was what was indicated as the result.

Is there any software I could use that might be better for my species? I just need to do this (very) quick and dirty (for now). I am interested to determine biological processes or functions of these DEGs I have. Here is a sublist of the DEGs, in case:

(I appreciate any help!!!!)

aug_v2a.00154.t1
aug_v2a.00307.t1
aug_v2a.00601.t1
aug_v2a.00603.t1
aug_v2a.00823.t1
aug_v2a.00990.t1
aug_v2a.01279.t1
aug_v2a.01408.t1
aug_v2a.01492.t1
aug_v2a.01496.t1
aug_v2a.01618.t1
aug_v2a.02238.t1
aug_v2a.02239.t1
aug_v2a.02459.t1
aug_v2a.03293.t1
aug_v2a.03318.t1
aug_v2a.03417.t1
aug_v2a.03732.t1
aug_v2a.03955.t1
aug_v2a.04041.t1
aug_v2a.04353.t1

Answer 1

[Devon Ryan]

I'm guessing that you're the same person that posted this thread on SEQanswers.

So those aren't gene names/symbols/IDs, they're just sequential identifiers given to the predictions (perhaps from PASA, which is one of the tools they used). What you're looking for is Blast2GO, or something like it, at least unless the team that produced that data has already blasted transcripts against another annotated genome to try and produce more useful gene names.

Answer 2

[cdsouthan]

Being ~ 0.8 billion years diverged from mammals A.digitifera won't (indeed should not) have any Official Gene Symbols via transative annotation

If you ran InterPro to GO on the ORFs (depending on thier quality) you might get something

(Just saw DR's answer on the same lines, JFTR, BLAST to GO may be eisier to pipeline than IntePro to GO)

Answer 3

[GreenDiamond]

Thank you so much both for your answers! I am very thankful for it.

I am working with Blast2Go right now. I have never used it before. But I am thinking I should follow protocol "Functional annotation in 3 steps: BLAST to find homologus sequences, MAPPING to retrieve GO terms and ANNOTATION to select reliable functions." ?

I have the GUI open on my computer. However, I am unable to load the sequences so that I can blast them. All I have is a .txt file of a list of genes in this format:

ug_v2a.18827.t1
aug_v2a.19274.t1
aug_v2a.19609.t1
aug_v2a.20118.t1
aug_v2a.20301.t1
aug_v2a.20307.t1

It will allow me input types .fasta, .dat, and .annot.

As I am only doing this for quick and dirty practice, is this somewhat reasonable Blast-->Map-->annotation, and if so, how do I input this gene list? Should I change the format first (and to what!?) so that Blast2Go can recognize it?

Thanks so much again!.....

[st.ph.n]

Blast2GO also needs interproscan results to be successful, otherwise you'll just get top species hits distributions. You can use the EMBOSS getorf that came with interproscan (or install independently) to get open reading frames of transcripts, and then put that file into interproscan. What did you put into DESeq? There's no FASTA for A. digitifera transcriptome? I know there exists a draft genome. You can also input, if you have fasta, into KEGG to simply get pathways.