Open Workflows (Lightning talk): Conquering confounds and covariates in neuroscientific analyses with an open, high quality library

[jsheunis]

Conquering confounds and covariates in neuroscientific analyses with an open, high quality library

By Pradeep Reddy Raamana, Rotman Research Institute, Baycrest Health Sciences, Toronto, Canada

• Theme: Open Workflows
• Format: Lightning talk

Abstract

Given the ever increasing complexity and sample sizes of the open datasets , there is clear need for deconfounding methods in various facets of neuroscientific analyses including predictive modeling. Recently this important topic has been getting increasingly more attention. However, there are still many challenges, including but not limited to lack of consensus on 1) what really constitutes a confound?, 2) when should we try to defoncound it? and 2) how do we properly assess their impact? etc. This calls for bridging a clearly unfilled need for a well-tested high-quality software library implementing the deconfounding methods as well as related tools to answer the aforementioned questions and open challenges. Towards this end, I built a python library called confounds, that is extensible and built for development with a community-first attitude following the best practices of open science. I would like to present its features, roadmap and encourage contributions from the deconfounding enthusiasts of all levels.

By conquering confounds, I mean methods and tools to

• visualize and establish the presence of confounds (e.g. quantifying confound-to-target relationships),
• offer solutions to handle them appropriately via correction or removal etc, and
• analyze the effect of the deconfounding methods in the processed data (e.g. ability to check if they worked at all, or if they introduced new or unwanted biases etc).

Useful Links

https://github.com/raamana/confounds https://crossinvalidation.com/2020/03/04/conquering-confounds-and-covariates-in-machine-learning/

Tagging @raamana

[raamana]

Yes, these details are correct.

[raamana]

I am unable to edit the top issue, so can someone edit the above making a note that second link has slides? Thanks.