different annotation for contigs in the same cluster

[junefo...@gmail.com]

Hello!

I wanted to consult regarding the result of my corset run. There is this cluster with more than 10,000 sub-clusters (Y) that showed different annotation for some sub-clusters. For example, Cluster123.15471 and Cluster123.18358 gave different nr results: leptin receptor gene and Heat shock factor 2 protein, respectively. I am planning to perform gene-level differential expression (using tximport), but this could introduce a large bias in DGE analysis. Is it not advisable to just consider a sub-cluster ID as a uniq cluster? I used salmon and corset with the default settings to generate the clusters.

Any suggestion will be appreciated. Thank you so much!

[Nadia Davidson]

Hello,

The "." is perhaps a little confusing here, but indeed the cluster ID corresponds to the full string, so Cluster123.15471 and Cluster123.18358 are different clusters. Corset should reports the counts (which can be used for DGE) for these separately.

Cheers,

Nadia.