is always necessary to do an intermediate index generations and 2nd pass alignment with STAR?

Question

[Lila M]

Hi everybody, I've just wondering if is totally necessary to do an intermediate index generations and 2nd pass alignment with STAR. Does anybody has information about when to skip it or not? I would like to know a bit more about this issue and to decide when is the moment to apply them or not.

Thank you very much in advance.

Best,

Answer 1

[arta]

Well according to documentation, in order to detect novel junction discovery, you need to apply 2-pass mode.

But, fortunately, if you use -- twopassmode, STAR will perform the 1st pass mapping, then it will automatically extract junctions, insert them into the genome index, and, finally, re-map all reads in the 2nd mapping pass (directly taken from documentation).

[Lila M]

thank you for this! But in that case, is still necessary to create the intermediate index? or Can I apply the --twopassMode in the 1st pass?

[Eric Lim]

The 2-pass mode isn't required to detect novel junctions. The 2nd step, however, does increase the mapping of detected novel junctions. Star 2-pass is most commonly used by combining all the sj files from the first pass. You can either use --sjdbFileChrStartEnd to provide a list of sj files (on-the-fly method), or reindex. If you choose to insert junctions to your genome using the on-the-fly method, you might need --limitSjdbInsertNsj when the total number of junctions are greater than 100k. Refer to the documentation quoted above for more details.

Edit: If you only have one sample or intent to perform 2-pass in a sample-by-sample basis, then --twopassMode will do.

[arta]

You can apply the --twopassMode in the first pass, at least i do.