Gene characteristics

[a-pankov]

I think you should try to account for gene characteristics. For example, longer genes might be expected to harbor more SNPs. These genes might also be the ones harboring more motifs for post-transcriptional regulation and might thus be studied more often. Also, take a look and see if the number isoforms per gene helps explain away any info associated with SNPs.

[cgreene]

We definitely thought about this. This analysis is more of a cautionary note for min(p) type approaches for network-based and other methods, so I think that the raw number of SNPs might be most appropriate for such approaches.

[dhimmel]

@a-pankov, these are great ideas. Unfortunately, they are outside of the scope of the current analysis.

However, the results are guaranteed to be interesting and high impact. I'll add a Zenodo DOI to the repo soon, so hopefully others will be able to use our current work as a starting point.