Computation time

[Elisa89m]

• spladder version: last (I guess)
• Python version: 3.9
• Operating System: Ubuntu

I installed spladder by the command pip install spladder under conda environment with python 3.9 since under an environment with python 2 it returned me dependence errors.

I'm trying to analyze the splicing variant in a mus musculus cell line, I downloaded the GTF reference (from ensembl).

I run the following command
spladder build --parallel 2 --output-txt-conf --output-gff3-conf -o ./Panc02_test_2 -b Panc02_input/uniqueTUMOR_sorted.bam -a ../Mus_musculus.GRCm38.102.chr_mod.gtf -c 3

The following files were generated in few hours:
genes_graph_conf3.merge_graphs.count.hdf5
genes_graph_conf3.merge_graphs.gene_exp.hdf5
genes_graph_conf3.merge_graphs.pickle
genes_graph_conf3.uniqueTUMOR_sorted.pickle

But the splicing variant files hdf5 were not generated even if I waited two days.
I run also the example test as following:
spladder build --parallel 2 --output-txt-conf --output-gff3-conf -o ./test_spladder -b input_spladder_test/testcase_events_1_sample1.bam -a input_spladder_test/testcase_events_spladder.gtf

and this was fine.

Similar to cell line test I performed a run on human sample by using the GTF reference downloaded from ensemble, but I gained the same problem.

Is there something that I wrong? Why the computation time is so long?

Thanks in advance for you availability.

Elisa

[akahles]

Dear Elisa,

sorry for the late reply. If your issue is still relevant, please send a ping and I will have a look. If you use --verbose in the output, it might give you a hint for where the program hangs.

Further, I recommend running with --sparse-bam, which summarises your bam files first into a format that SplAdder can interact with more efficiently.

Lastly, try using the latest release. Depending on which aligner you used, you might have been affected by a bug that occurred when parsing CIGAR strings that contain = or X characters.

Best,

Andre