README.md

kinase-robot

GoodDocData -- A Template for Simple and Clear Documentation of Hackathon Analyses!

adapted from NCBI-Hackathons/GoodDoc with some tweaks for analysis-driven projects

instructions in italics can be deleted as sections are filled in

most fields are optional, Conclusion and Important Resources are required

Please cite our work -- here is the ICMJE Standard Citation:

...and a link to the DOI: You can make a free DOI with zenodo, synapse, figshare, or other resources

Awesome Logo (if applicable)

Website (if applicable)

Abstract : Summarize everything in a few sentences.

Introduction : What's the problem? Why should we solve it?

Methods : How did we go about solving it?

Results : What did we observe? Figures are great!

Conclusion/Discussion:

Please make sure you address ALL of the following:

1. What additional data would you like to have

• Additional kinome data from all drugs and compounds in the world
• Toxicity data based on kinases inhibited
• Pathway interactions from kinases to others, up and down regulating (Kegg?)
• More Kinome screening data from the tumors we're trying to identify drugs for (to cancel out error rates)

2. What are the next rational steps?

• Adding to the pipeline
• Implementing multiple pipeline components, such as feedback loops from drug screending data (both positive and negative)
• Ability to use Kinase expressions from a single tumor for personalized drug predictions
• Add toxicity predictions
• Add combination therapy predictions
• Add RNA predictions

3. What additional tools or pipelines will be needed for those steps?

• Toxicity prediction module
• Pathway extentions, to amend the kinase data with downstream regulating factors
• Implement module for feedback loop on durg screening data

4. What skills would additional collaborators ideally have?

• Math
• Biology, drug screening

Reproduction: How to reproduce the findings!

1. Scrape the Synodos Kinase profiling into database using loadKinese function
2. Download all Lincs data using the downloadLincs function to local machine
3. Populate kinase inhibition molecule data from Lincs data with function loadLincs
4. Use scoreKinase function to rank the important kinases for the relevant tumor sample, eg: baselines=Syn1_SF,Syn2_SF and alternatives=Syn5_SF,Syn6_SF,Syn8,Syn10,Syn11
5. Calculate the best drugs using the scoreDrugs function

Important Resources : primary data, github repository, Synapse project, dockerfile link etc.