Repository navigation
Web Server For Displaying Omics Samples for Selected Tissue #5
Description
Group
Arjun Krishnan
Contact info
Arjun Krishnan, PI, ARJUN.KRISHNAN@CUANSCHUTZ.EDU
Chris Mancuso, Lead Contact, Christopher.a.mancuso@cuanschutz.edu
Links to code
https://github.com/krishnanlab/meta2onto_scripts
https://github.com/krishnanlab/meta2onto_webserver
https://github.com/krishnanlab/meta2onto_api
I added Vince and Faisal. Let me know if I should anyone else.
Workflow
We have been using git to everything. I think most of this was done on Microsoft Azure. The repos describe the software which I'm not that familiar with myself but the main pieces are Azure, React, ElasticSearch, Flask and PostgreSQL. However, any of those can be changed to whatever works best.
I have about 50 MB of other files that are needed and can email them when is needed as well.
Work description
There is no PHI as all of this is public data.
The over goal is to build a web server where someone can type in a tissue name and web server will fetch the samples that are the most probable to be that tissue type (probabilities are pre-generated through a method published in the group). The data returned will be metadata description of the omics samples, so some structured texted but also unstructured text. Additional feature will be highlighting key works (determined in advance per sample by the NLP method), allowing users to save their lists and download the sample IDs.
The first phase would be to look over code from MSU and see what can be reused. The second phase is getting the web server up and running on a small set of data. The third phase is to add extra features to the UI. Phase four is testing with all the data (i.e. finding relevant samples from millions of IDs). Phase 5 is deploying the web server publicly. The last phase is to write a brief manuscript on the web server.
Long term maintenance will be needed.
Timeline
A hard deadline is Summer 2024 as I believe this is when the grant will end.
Funding
This will go under Arjun's NSF funding for the Career award.
Activity