Clarification on RSEM, dedup and STAR 2-pass

[lianov]

Hello,

Just a couple of points on RSEM utilization in your pipeline:

1. In reference to the pipeline described here I don't see it being mentioned that one of the dependencies is an RSEM reference shown by REF=/srv/scratch/restricted/rare_diseases/data/quantification/rsem/rsem_reference/hg19_RSEM from rsem_calculate_expression_raredisease.sh

2. In the documentation it would be helpful to state that rsem-calculate-expression is the specific command that is being utilized from RSEM (as it contains several).

Questions:

1. Did you/your team considered using STAR 2-pass for novel junctions? Wondering if that may cause any differences in the context of rare disease.

2. What is the reason for deduplication here for RNA-seq? Normally, RNA-seq reads are not deduplicated as it's difficult to differentiate PCR duplicates from biological duplication of mRNA (RNA abundance). My guess, is that this was a conservative approach as there are several other filters in this analysis as well, but would love to see a more in-depth explanation.

Thank you!