Add genes to "ion channel activity" annotation

[jastrother]

I'm in the process of analyzing some zebrafish neuron transcriptome data. I made a list of genes with ion channel activity based on zfin and a review of the literature. I just compared against the amigo "ion channel activity" annotation, and of course I missed some but there also seem to be a few genes missing from amigo. I'm new to GO and amigo, so I'm not sure what to do with this information. Here were the genes that appeared to be missing from the annotation:

Missing: grin2ba :: glutamate receptor, ionotropic, N-methyl D-aspartate 2B, genome duplicate a Missing: gabra2b :: gamma-aminobutyric acid type A receptor subunit alpha2b Missing: gabrb3 :: gamma-aminobutyric acid type A receptor subunit beta3 Missing: gabrg2 :: gamma-aminobutyric acid type A receptor subunit gamma2 Missing: gabrg3 :: gamma-aminobutyric acid type A receptor subunit gamma3 Missing: gabrr2b :: gamma-aminobutyric acid type A receptor subunit rho2b Missing: chrna7b :: cholinergic receptor, nicotinic, alpha 7b Missing: chrna9b :: cholinergic receptor, nicotinic, alpha 9b Missing: chrna10b :: cholinergic receptor, nicotinic, alpha 10b Missing: chrnb2bl :: cholinergic receptor, nicotinic, beta 2b like Missing: kcnk4b :: potassium channel, subfamily K, member 4b Missing: kcnk12 :: potassium channel, subfamily K, member 12 Missing: kcnt2 :: potassium channel, subfamily T, member 2 Missing: hcn2a :: hyperpolarization activated cyclic nucleotide-gated potassium channel 2a Missing: clca1 :: chloride channel accessory 1 Missing: clca5.1 :: chloride channel accessory 5, tandem duplicate 1 Missing: clca5.2 :: chloride channel accessory 5, tandem duplicate 2 Missing: ano1b :: anoctamin 1, calcium activated chloride channel b Missing: ano7 :: anoctamin 7 Missing: ano8a :: anoctamin 8a Missing: ano8b :: anoctamin 8b Missing: ano9a :: anoctamin 9a Missing: ano10a :: anoctamin 10a Missing: ano10b :: anoctamin 10b Missing: scn4ba :: sodium channel, voltage-gated, type IV, beta a Missing: scn4bb :: sodium channel, voltage-gated, type IV, beta b Missing: kcna3a :: potassium voltage-gated channel, shaker-related subfamily, member 3a Missing: kcna3b :: potassium voltage-gated channel, shaker-related subfamily, member 3b Missing: kcna5 :: potassium voltage-gated channel, shaker-related subfamily, member 5 Missing: kcna6b :: potassium voltage-gated channel, shaker-related, subfamily, member 6 b Missing: kcna10a :: potassium voltage-gated channel, shaker-related subfamily, member 10a Missing: kcna10b :: potassium voltage-gated channel, shaker-related subfamily, member 10b Missing: kcnab3 :: potassium voltage-gated channel, shaker-related subfamily, beta member 3 Missing: kcne1 :: potassium voltage-gated channel, Isk-related family, member 1 Missing: kcnk3l :: potassium channel, subfamily K, member 3, like Missing: cngb1b :: cyclic nucleotide gated channel subunit beta 1b Missing: piezo2a.1 :: piezo-type mechanosensitive ion channel component 2a, tandem duplicate 1 Missing: piezo2b :: piezo-type mechanosensitive ion channel component 2b

[sabrinatoro]

@pgaudet I went through the list of genes above. I cannot find any evidence in the literature that these genes should be annotated to "ion channel activity". Here are the reasons: 1) no experiment was made to show that these genes have this activity. 2) Authors assume that these genes have the "ion channel activity" because they are phylogenetically related (and in some case, share sequence similarity) with the human orthologs, However, when I looked in Amigo, the annotations of the human orthologs were not based on experimental evidence, and therefore, I cannot make an annotation using an ISS or ISO as evidence (since what would be in the 'with' field does not have experimental evidence).

Maybe these should/could be annotated via PAINT?

Please advice. Thank you!

[pgaudet]

@sabrinatoro Can you provide the ZFIN Ids for those, so I can look for them in PAINT ?

Or the PANTHER families - you can query here: http://www.pantherdb.org/

Thanks, Pascale

[sabrinatoro]

@pgaudet The IDs and their mapping to Panther can be found here: https://docs.google.com/spreadsheets/d/1rwljd89SIxvxx0JrsZydRgrJlU1knQMATqKIIModyFk/edit?usp=sharing

Please note that some of the genes/IDs couldn't be mapped in panther.

Thank you, Sabrina

[marcfeuermann]

This is an annotation issue, not a PAINt issue.

[RLovering]

I guess many of the human annotations are TAS and NAS annotations or IEAs please do not remove the human annotations due to a lack of expt support for the annotation of these essential channels

[pgaudet]

We were not planning to remove any annotations

[ValWood]

Since this is a Zfin annotation issue, should it be submitted to the Zfin helpdesk. @jastrother do you still need this, the ticket is quite old?

One option if they cannot be PAINTED, or we don't have experimental data to infer from would be to improve the InterPRO mappings. Or for @jastrother to supply the IDs of the annotated orthologs to make ISS annotations from.

Since @sabrinatoro is no longer at Zfin, I'll assign to @doughowe and @sramachand to decide how to proceed.

[ValWood]

@pgaudet I don't seem to be able to assign this ticket?

[jastrother]

We were able to work around the issue by making our own internal database, which supplements the GO information based on some manual curation. I thought it would be good if the public information was also updated, but I don't really understand the process for making these changes. As you mentioned, the ticket is quite old, and we already have a working solution, so I'm not opposed to closing it at this point.

Best, Jim

[doughowe]

@sramachand It looks to me like the manual annotations we can make have already been done (per Sabrina's review of associated literature). Any automated annotations that could come via PAINT or translation files would happen if those resources included them. So, I don't think there is anything we can do here, so closing seems appropriate.

[ValWood]

@jastrother

. I thought it would be good if the public information was also updated, but I don't really understand the process for making these changes.

This tracker is sometimes used to suggest annotations but its main purpose is for correcting misannotations. The best way to get annotations added is to contact the annotation provider directly, which would be the individual Model organims database for the major model species and UniPRot for human and everything else.