biological meaning /summary for similar sets of GO terms

Question

[lessismore]

Dear all, starting from limited sets of GO terms (expressed with GO IDs or GO descriptions), is there a way to find shared ancestors to better describe the summarised biological meaning of those GO term sets?

Thanks

[Istvan Albert]

I am developing a software for this purpose it is is not quite done yet, but you could be an evaluator if you are interested,

the software has both command line interface and it can also visualize via a web browser

[lessismore]

Sure, it would be great

[Yogi]

I would also be interested in this Istvan Albert !

Can you post in here when you actually finish the project?

Would be glad to also post about it on my LinkedIn.

Answer 1

[Istvan Albert]

I am developing a tool I call Genescape that, among many others offers the functionality you are after.

https://github.com/ialbert/genescape-central

the tool is under heavy development at this time, but I think it can already do what you are after. Basically if you have a list of GO terms or a list genes, you can visualize a tree or a subtree of it:

Basically from a list like this:

GO:0005488
GO:0005515
GO:0048029
GO:0005537
GO:0003824
GO:0030246
GO:0036094
GO:0016787
GO:0043167
GO:0016788
GO:0140375
GO:0046872

It can draw something like this below (green nodes are terms in your list, blue nodes are terminal nodes (also in your list)) white nodes are intermediate nodes

Answer 2

[geneontologyhelp]

This is already available through the QuickGO browser from EBI. The Basket feature allows you to view an ancestor chart:

From the Basket you can look at or download a complete ancestor chart, complete with Slims and other custom options. The boxes in the chart are links back to the QuickGO term page.

[Istvan Albert]

it is good to know, and thanks for the update,

though I will say the feature is so quite hidden and obscure - I have been teaching bioinformatics for over a decade yet I was unaware of it.

Most importantly, the genescape tool has many other features, and the tree drawing is just one of the visualizations it provides.

The tool tries to offer a reproducible and automatable approach to whittle down huge and complex functional trees into more manageable ones. One can select only branches that deal with regulations etc.

When one deals with a list of genes the ontology tree can be gigantic, I am exploring ways to focus on various subparts of the tree and see how well those are supported.