support CCS data

Question

support CCS data

Arkarachai opened this issue 6 years ago · comments

Arkarachai Fungtammasan commented 6 years ago

Would it be possible to support CCS data as well? I think you would need to make an option to pass in preset mapping parameters keyword to Minimap in SDA.1.snakemake.py and SDA.2.snakemake.py, but I'm not sure if any other part of the program needs to be changed.

Mitchell Robert Vollger · Answer 1 · Fri Apr 26 2019 04:24:46 GMT+0800 (China Standard Time)

Hi Arkarachai, I am working on updates to allow for CCS data that I have not yet pushed.

I will update you when I have pushed them. Thought I will say in my preliminary analysis is that CLR data does better than HiFI with current parameters because the length of the read is very important in linking together distant PSVs and HiFi is quite a bit shorter.

Best,
Mitchell

Arkarachai Fungtammasan · Answer 2 · Fri Apr 26 2019 04:29:48 GMT+0800 (China Standard Time)

Sounds good. Regarding your update. Would there be any change beside preset parameter for minimap? Also, do you have the rough timeline of the new release?

Mitchell Robert Vollger · Answer 3 · Fri Apr 26 2019 04:42:09 GMT+0800 (China Standard Time)

I have a pressing project at the moment that will be done at the beginning of next week. So I hope sometime next week.

Just an FYI minimap2 will align CCS reads pretty well with just the map-pb since it is forced to do dynamic programming alignments so if it is urgent you could just try it and see what happens.

The other changes I would like to make is that there is a statistical test for the fraction of reads that have to have PSVs in order for two PSVs to be linked that needs to be changed to reflect the error rate of HiFi instead of CLR.

Mitchell Robert Vollger · Answer 4 · Tue Nov 05 2019 07:32:43 GMT+0800 (China Standard Time)

It took much longer than I hoped to get back to this, but it is now done as of #10. Fortunately, this update comes with changes that should make SDA a bit easier to run.