error MarkerScanner.pl

Question

[Kárita]

I'm trying to run amphora2 but every time I get the following error:

Can't open /Marker/marker.list at MarkerScanner.pl line 89.

Would anyone know how to solve it?

[Ram]

What is the exact command you're running, and what kind of machine are you running it on? What OS are you using? Give us as much detail as you can.

[Kárita]

I'm using a Google server with 32 cores and 128g of memory. Linux operating system.

The command used is the following:

perl MarkerScanner.pl -Bacteria -DNA TestData/contigs.fasta

however any within the Amphora do not work

which aims to: 1. Marker identification Use MarkerScanner.pl to identify bacterial and/or archaeal marker sequences. Given a sequence file, this program will identify markers from the input sequences and generate a protein fasta file for each marker gene in your working directory. For example, rpoB.pep, rpsJ.pep. When DNA input sequences are used, this program first identifies ORFs longer than 100 bp in all six reading frames, then scans the translated peptide sequences for the phylogenetic markers.

it uses the following software as support: AMPHORA2 depends on several external programs.

1. HMMER3 (http://hmmer.janelia.org/). Required for marker identification, sequence alignment and trimming. Earlier versions of HMMER will not work!
2. RAxML version 7.3.0 or later (https://github.com/stamatak/standard-RAxML/downloads). Required for phylotyping.
3. Bioperl 1.5.2 or later (http://www.bioperl.org/wiki/Getting_BioPerl).
4. EMBOSS (http://emboss.sourceforge.net/download/). The 'getorf' program of the EMBOSS package is required only if you analyze DNA sequences using AMPHORA2

All are installed and tested for functionality.

[Ram]

Disclaimer: I know nothing about these tools but I am good at perl and shell scripts.

I looked into the source code and see the following at MarkerScanner.pl line 89:

open (IN, "$AMPHORA_home/Marker/marker.list") || die "Can't open $AMPHORA_home/Marker/marker.list";

This references $AMPHORA_home, which in line 25 is taken from the AMPHORA2_home environment variable. Given your error message has no value for $AMPHORA_home in "Can't open $AMPHORA_home/Marker/marker.list", I think your environment variable is not set and that is contributing to the problem.

[Kárita]

Okay, I managed to solve this, then I got a new error:

Find and extract open reading frames (ORFs)

Error: File existence/permissions problem in trying to open HMM file /home/bryoantar2023/software/Amphora_2/AMPHORA2/MarkerBacteria.markers.hmm.
HMM file /home/bryoantar2023/software/Amphora_2/AMPHORA2/MarkerBacteria.markers.hmm not found (nor an .h3m binary of it)

Can't open 26019.hmmsearch at MarkerScanner.pl line 100.

[Ram]

I think I understand why this is happening but I am not a 100% sure. This seems to be a weird unstated convention used within the code. The output from the hmmsearch operation on line 81 saves results to a file with the current process ID as the prefix. If your current PID does not match the PID when the hmmsearch was run, line 100 will fail. Ideally, these out prefixes should either be static or user-defined, not a random per-run value like it's done here.

Can you try removing ALL files generated by this script and re-run from scratch?

[Kárita]

Do you suggest I delete the program and install it again?

[Ram]

No, I'm recommending you delete the files generated by previous runs and try again, or simply create a new dir, link to your input files and run the script there.