Closed cmandreani closed 2 years ago
nselem
commented
2 years ago Please could you send me a sample with ~10 genomes to see if I can reproduce the error. Can you show me the header of some input file? Also, could you tell me if the sample that worked ok for CORASON?
cmandreani
commented
2 years ago Hi, thanks for answering.
I am not using genomes, as it is explained here that:
CORASON finds variation in the genomic vicinity of a reference cluster. To this end, CORASON can explore either BGCs predicted by antiSMASH or complete genomes. Results of this approaches will be slightly different.
And in the Streptomyces example, running is shown as:
~/bin/run_corason TauD.fasta gbks gbks/JMGX01000001.1.cluster003.gbk -g
~/bin/run_corason TauD.fasta genomes genomes/JOBW01.gbk -g
I'll send to you the input files (queryGene.faa and refBgc.gbk) and the 10 genomes vía email.
Cheers.
nselem
commented
2 years ago Just for the sake of other users: This issue is similar to this one: https://github.com/nselem/evomining/issues/3 The problem is the use of muscle in ubuntu for windows. And corason stand-alone is better for exploring genomes while searching for BGCs, or BGCs variants that may have not been annotated by antiSMASH. If you already have your Genbank files of your desired BGCs there is no need to use CORASON standalone tool because big-scape already have a corason version integrated that allows visualizing the BGCs phylogenetic trees , and it is nod needed for exploring because bigscape already
cmandreani
commented
2 years ago I've noticed that CORASON's third issue is also solved with this answer.
Cheers and thanks again.
Ubuntu 20.04 (WSL)
Hi, I've installed BiG-SCAPE and CORASON successfully.
When running BiG-SCAPE it appears to be no problem as I can visualize with no troubles the index.html file.
However, when running:
~/bin/run_corason query.fasta gbks_dir/ gbks_dir/ref_BGC.gbk -gthe output folder is generated with several written files such as 72 "JobID.input", 8 "ClusterN", "Concatenados.faa", "TempConcatendados.faa", "Frecuency", "query.fasta.BLAST", "query.fasta.parser", "query.fasta_PrincipalHits", "query.fasta_Report" and others, plus 3 folders (CORASON, GBK, MINI). But "query.fasta_tree.svg" is blank (unable to open with firefox and weights 0 KB).
I noticed that the "Corason_Rast.IDS" file generated didn´t collapsed BGCs into their genomes correctly (there are 1911 BGCs from 192 genomes, but their genbank files are named i.e.: genomeY_ctgZ_region001); for which i replaced the second column with the "genomeY" annotations, keeping the same architecture than the original (JobID/GenomeID/OrganismName) and ran:
~/bin/run_corason query.fasta gbks_dir/ gbks_dir/ref_BGC.gbk -g --rast_ids CorasonRast_indexMod.IDSThe following result presented no change despite the correction of the Corason_Rast.IDS file:
I've selected several BGCs among the dataset, with or without matches in MiBIG, varying the query genes from core to adjacent and belonging to the reference BGC or from any other, but with no success.
I tried also lowering down the cutoff to 1E-5 and extending the cluster radio up to 70 with no results.
Any ideas where may I be failing?
Thanks.