Skip to content

Latest commit

 

History

History
49 lines (40 loc) · 3 KB

idea-page.md

File metadata and controls

49 lines (40 loc) · 3 KB
Raw

For the start of this project, this will be our idea page. People can post ideas here that could lead to projects. Additionally, if people have journal articles or software tools that they think could be useful/worthwhile to discuss, this could be a good space to put those during our discussion. For organization, you should create a section with your name/identifier, and underneath that put whatever you want to dicuss. This will allow others to follow your ideas and thought processes as they develop.

Points from Slack discussion

  1. We want to be able to support as many genomes/organisms as possible (eukaryotes?)
  2. We want a localhost-based dashboard with interactive visualisation
  3. We want to be able to at least analyse RNA-seq data
  4. Combining this data with things like gene expression/COG/etc is also something mentioned multiple times
  5. TO DISCUSS: data will be read in using either DB's, flatfiles or both
  6. The analysis should be done either interactively using the dashboard, or more statically by calling the functions through a library

Suggestions:

Matt Moss

  • I don't have any concrete ideas yet, but I think something to do with the cancer genome atlas database could be a good starting point. There's a lot of raw/partially processed next-gen sequencing data that for sure still has interesting conclusions buried within it. I'd also say that doing something involving bam processing could be worthwhile, but I don't know what

Mark van der Sman

  • An analytical pipeline for genomics/transcriptomics accompanied with a dashboard might be awesome! No real application that I can think of except for something that lets you delve into the data and attach multiple sorts of datasets.
  • After some discussion on Slack:
    1. A combination of genomics and transcriptomics seems preferred so far.
    2. 3 general parts: DB, general middleware with breakout ML group (Deep Learning?), Visualisation.
    3. Infrastructure/computing side MIGHT be possible through the cloud provider a member works at.
    4. sc-RNA seq tools seem to be in demand.

clintval

Bioinformatics technical blogpost and tutorial aggregator.

  • A lot of educational resources exists by those in our field that are experts in certain domains. Many of them write posts on best practices, new tools, new findings. It would be great to have this content ready-to-server in one easy to navigate place.
    • Content is provided by user submission
    • Blog posts are tagged with domain of resource or tools used
    • A simple click to vote could display more relevant or useful metered information

A complete Python module for dataviz in genomics

  • R has packages like it like GenoPlotR and Sushi, etc.
  • Would provide a consistent interactive Python API
  • Could be built to abstract away plotting library (matplotlib / bokeh / plotly)
    • Similar to how graphics packages allow you to choose a back-end
    • Would require some thoughtful engineering
  • Crowd favorite plots can be built first (topics included Sushi-plots, Circos wrappers)