4 Comments
User's avatar
Muralidharan Kaushik's avatar

Hello Evan, thanks for the article! I have been deeply thinking along similar lines for some time and have found many insightful ideas from your writings. Do you have any thoughts and opinions regarding usage of simulations for biological analysis? I would really like to know if you do!!

Evan Peikon's avatar

Hi Muralidharan, thanks for your comment. I've previously written about modeling and simulating branch pathway systems, which you can find here (https://sequenceanddestroy.substack.com/p/computational-modeling-of-branched?utm_source=publication-search). In general, I don't do a lot of simulation work; not because I don't find it useful, but because it's not highly relevant for my specific research interests. To the extent that I do use simulations, it's typically for simulating the effects of inhibiting specific proteins on a network to identify proteins whose removal results in the largest perturbations.

gmdbioinformatics's avatar

Thank you for the article, very insightful!

It's true that, when you look at different types of data, results are often different.

Many people talk about multi-omics approaches. However, in this case, I think that the three signals you have observed may have been diluted, and none of them observed.

User's avatar
Comment removed
Dec 1
Comment removed
Evan Peikon's avatar

Agreed; I definitely think this type of approach can benefit from more formally incorporating PK constraints. At the moment, my approach has been to use this type of framework to generate a list of candidates, after which I'll manually look at the target's druggability, PK profile, literature supporting it's activation/inhibition in similar contexts (or not), etc. But, having a way to codify that process would be very useful, as would the ability to dynamically adjust weights as you mentioned. Thanks for the input!