lh3 / minigraph

Sequence-to-graph mapper and graph generator
https://lh3.github.io/minigraph
MIT License
419 stars 38 forks source link

Illegal instruction #110

Open Mar10L opened 8 months ago

Mar10L commented 8 months ago

Hi. I'm trying to generate a graph from 3 fasta files (reference + 2 samples).

If I run the command with the test files, all is going well. I tried the command with my dataset (1 chromosome for each fasta file ~26Mbp):

minigraph -xggs -t20 reference_chr01.fasta sample1_chr01.fasta sample2_chr01.fasta > out.gfa

but i get this error: _*[M::main::0.0021.68] loaded the graph from "reference_chr01.fasta" [M::mg_index::0.0182.09] indexed the graph [M::mg_opt_update::0.0201.95] occ_max1=26; lc_max_occ=2 [M::ggen_map::0.0221.82] loaded file "sample1_chr01.fasta" [M::ggen_map::0.0771.25] mapped 1 sequence(s) to the graph Illegal instruction**_

I noticed that, giving just the first 59940bp, the command is working. Instead, by giving one more row in each fasta (60bp more), the program is not working anymore.

P.S. I get the same error by using the -c option, as suggested in the usage example, even with the test dataset.

Best Mario

colindaven commented 8 months ago

You're probably adding special characters to the file. Try checking with

dos2unix

dos2unix *.fasta

show non-printing chars

cat -e file.txt

grep non-printing characters

grep -P -n "[\x00-\x1F\x7F-\xFF]"

vim view special characters

Re: How to view special characters in text files ?
well, VI has some options for this

:set list

lh3 commented 7 months ago

Are you using precompiled binary or are you compiling from source on your machine?