Issue running MethylDackel extract in parallel mode using minConversionEfficiency

[blue-moon22]

Hi there,

I'm having an issue running MethylDackel extract in parallel mode using minConversionEfficiency for some samples and not all.

For those samples that fail (call them Sample 1 and Sample 2), I'm running it in a Nextflow process within a Docker container (with MethylDackel version 0.6.1 installed with conda):

MethylDackel extract -@ 8 --mergeContext --minConversionEfficiency 0.9 /tmp/index/$index_file $bam -o ${prefix}.bedGraph

For Sample 1, this always fails. For Sample 2, it mostly succeeds, but sometimes fails.

When I am running with a single cpu -@ 1 or when I run it using --minConversionEfficiency 0.0 all succeed.

Backtrace of the core dump files consistently give me the following:

[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Core was generated by `MethylDackel extract -@ 8 --mergeContext --minConversionEfficiency 0.9 /tmp/ind'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x0000555c4c2704c9 in computeConversionEfficiency ()
[Current thread is 1 (Thread 0x7f40861fc700 (LWP 1815))]
(gdb) bt
#0  0x0000555c4c2704c9 in computeConversionEfficiency ()
#1  0x0000555c4c27077c in filter_func ()
#2  0x00007f408f7dde63 in bam_plp64_auto () from /opt/conda/envs/mitrabio/bin/../lib/libhts.so.3
#3  0x00007f408f7de40b in bam_mplp64_auto () from /opt/conda/envs/mitrabio/bin/../lib/libhts.so.3
#4  0x0000555c4c275ed3 in extractCalls ()
#5  0x00007f408f896fa3 in start_thread (arg=<optimized out>) at pthread_create.c:486
#6  0x00007f408f6074cf in __libc_ifunc_impl_list (name=<optimized out>, array=0x7f40861fc700, max=1)
    at ../sysdeps/x86_64/multiarch/ifunc-impl-list.c:357
#7  0x0000000000000000 in ?? ()

I hope this can be fixed soon as running this using 1 CPU takes a while. Thank you for your time!