timbitz/isosceles

The Isoforms from Single-Cell; Long-read Expression Suite

isosceles's Issues

find_isoswitch - from wf-epi2me single-cell transcript matrix

Hello,

thank you for the isosceles package. I was wondering if it was possible to work from nanopore wf-epi2me single-cell pipeline results with isosceles ?

special emphasis about "find_isoswitch" function that would be great if I didn't need "the Isosceles transcript" in order to make it work.

Best regards

ncpu doesn't seem to work

Hello, I tried to run scrna analysis using your tool by trying the multiprocessing functionality, i.e. adjusting the ncpu parameter. However, when I check my run using the top command, it was capped at 100% cpu usage. What could be the potential problem?

prepare_bam_transcripts.R duplication in hash id

Hi, thanks for the tool and for the great documentation.

In the prepare_bam_transcripts.R line 118, the code asked to verify that the hash id generated are unique. However, the hash id generated is not unique using my dataset. After looking into the code, I realise this was because some of the tx_intron_positions are the same despite having different tx_position (different chromosome). I believe the reason for this was the strands are undetermined, causing the overlapping, and hence generating the same hash_id based on the same tx_intron positions.

I replaced make_hash(tx_intron_positions) by make_hash(paste(tx_position, tx_intron_positions)) in line 112 to give an unique hash_id, do you think this make sense ?

Remark: I also replaced make_hash(.data$intron_positions) into make_hash(paste(.data$position, .data$intron_positions)) for consistency.

Alternatives for DEXSeq DEU test

Hi,
First, thanks for a careful development of this package.
I have been trying several aspects of Isosceles, however, I usually get stuck when I need to run DEXSeq. My dataset is way too big and runs days, sometimes not even managing to get the estimateDispersions step finished after a week. I have tried parallelized it with more cpus (12) and filtering in many ways my dataset to reduce the number of events to test but I still get to run DEXSeq for days. For instance, I ran the window approach, with 1500 events and 480 window columns and it took 2 days to finish.
I read that DEXSeq is not recommended for long datasets, but rather approaches like satuRn or limma. I understand that satuRn only performs DTU test, which is not exactly what is intended with DEXSeq (with testforDEU). Do you have a suggestion to deal with bigger datasets (more than 1000 cells and more than 1000 events)? Could it perhaps edgeR's diffSpliceDGE be adapted for this step?

Thanks,
Mariela