This is a running list of proposed features at WormBase
- 1 New displays, analysis, and data integration
- 2 Tools
- 3 General UI
- 4 Back end
New displays, analysis, and data integration
We need broader integration of human diseases. This is a benefit both to people who use C. elegans as a model system (most grants at least true to touch on the relevance of given projects with human disease) as well as those looking to the system to help understand orthologous genes implicated in disease.
- should be possible to search by disease IDs and common names
- genes with human orthologs implicated in disease should be clearly noted, probably in the Overview widget.
- Should link back to OMIM
- OMIM IDs in database; need brief descriptions?
- Is there an OMIM web service we can consume?
We don't too much with ontology other than display a list of associations for a given object. What else can we do? For example, on the gene page, we could expand the ontology section to show genes with similar patterns of enrichment. We might also consider integrate a tool like AmiGO to make it easier to browse ontologies.
- show other genes with similar (aggregate) gene ontology
- See GeneMania; algorithm for finding neighbors of genes (such as shared GO annotations)
- Has a nice API; catch an image
We list homologs and orthologs of genes but provide no alignments. These should be readily available from the gene page.
- (pre-generate) alignments with orthologs via BLASTX
- display in-line on the homology page or perhaps in lightbox
- when available, overlay variations to help identify conserved and critical motifs
- get them from the GO reference genome project.
- Predownload all the alignments
The ability to do forward genetics is one of the great strengths of C. elegans. We don't do enough with alleles identified in such screens. Can we gain additional information by looking at alleles in aggregate?
- Try to create an allelic series (organizing alleles from weakest to strongest) by looking at the type of allele (missense, nonsense, etc) and phenotype. Display on the gene page.
- Null alleles are of particular interest. Call these out in listings with special formatting.
- Map molecularly characterized variations against protein domains and cross reference with phenotypes to infer domain/protein function.
A particularly active area of research focuses on molecular evolution. Part of this research entails collecting SNPs from geographically distributed strains. We need better displays that allow exploration of this data. Marie-Ann has provided many suggestions. Here are a few more.
- A map-based display that lets users dynamically see which strains are carrying specific markers.
- Genome browser tracks showing the frequency of given SNPs in wild isolates.
Micro/Macro Synteny; whole genome alignment
We have nearly 20 species available through WormBase right now. Aside from individual gene pages and genome browsers, there is very little that one can do with this information.
- Regional nucleotide level alignments, shown as a small snippet, and presented on the gene page.
Pathways and Processes
As more and more labs turn to next generation sequencing, we need to start thinking about ways to facilitate use and collection of this data. We'll assume that end users will have already analyzed their data. Do we want to collect it? Should we provide a mechanism that makes it easy for people to, say, go from sequencing runs to identifying mutations?
- Sequence repository for C. elegans sequence data?
- Supplementary project with new funding
- via cytoscape web
- v3 in development; HTML5 based
Microattribution and Nanopublications
Microattribution and nanopublications are the wave of the future. We need to think about how to provide suitable functionality to allow people to contribute data and fix annotations as well as to receive credit.
- What can we build NOW as a proof of concept using WBIDs and our commenting system? For example, perhaps we could make the Brief Description editable on the website and list individuals contributions on their Person page. If done in a specific way, we could actually move much meta information on objects OUT of the acedb database into a more amenable database/interface structure.
Although we've made the "what's popular" on WormBase opt-in, I'd still like to see application of browsing history to try and elucidate previously unknown relationships between entities. For example, can we find previously undescribed gene interactions by looking at browsing history?
Public and private tags.
Currently we have GBrowse configuration in two places, as part of the website-gbrowse-support git repository and integrated into the new web app.
We need to maintain the website-gbrowse-support repository for the classic site. Going forward, I'd like to keep all config as a part of our app. For now, we need to make sure that any changes made to the classic site "website-gbrowse-support" repository also find their way into the configuration that is located in the new webapp.