heathsc
commented
4 years ago Hi Andy,
Could you try updating to the latest version from the master branch (including the sub-repository for bscall)? If you could then remove any editing .gemBS directory in your working directory, and then re-run the gemBS prepare and gemBS index steps. Let me know if this does not resolve the problem.
Thanks, Simon
p.s., the documentation is not completely up to date with respect to the various index files that are generated. This is something that is on the todo list!
On Wed, Mar 11, 2020 at 1:23 AM Andy Wang notifications@github.com wrote:
Hi GemBS team,
Thanks for designing gemBS this cool methylation calling tool. As it claims to perform better than bis-snp, our lab wants to do a snp calling to see how many SNPs we can get (right now, just test for one sample). We have computational nodes with gemBS 3.0.0 and 3.5.0 version.
I tried both versions of gemBS and they both worked well until the index step. However, for the next step "gemBS call" (I already hold the .bam files, so wanna skip the "map" step), they return different errors:
--- For gemBS 3.5.0, it returns:
$ gemBS call gemBS reference indexes/Homo_sapiens_assembly38.gemBS.ref not found. Run 'gemBS index' or correct configuration file and rerun
This makes me a little bit confused. As I checked gemBS's manual guide, the index step will not create a file with .gemBS.ref extension. Anyhow, then I tried to run "gemBS index" again and got this:
$ gemBS index 2020-03-10 16:29:16,026 ERROR: Process '/usr/lib/python3.6/site-packages/gemBS/bin/samtools' finished with 1 2020-03-10 16:29:16,026 ERROR: [faidx] Could not build fai index indexes/Homo_sapiens_assembly38.gemBS.ref.fai ValueError: Error while making faidx index of gemBS reference
I am not sure on what happened, but I guess there may be an error when I update the cluster node, which causes gemBS failing to load related related packages/dependencies. This needs to be double checked. If you know the reason, I'll appreciate a lot.
--- For gemBS 3.0.0, it returns:
$ gemBS -j test.json call Traceback (most recent call last): File "/usr/bin/gemBS", line 13, in load_entry_point('gemBS==3.0.0', 'console_scripts', 'gemBS')() File "/usr/lib/python3.4/site-packages/gemBS/commands.py", line 157, in gemBS_main instances[args.command].run(args) File "/usr/lib/python3.4/site-packages/gemBS/production.py", line 906, in run if self.conversion != None and self.conversion.lower() == "auto" and not args.concat: AttributeError: 'list' object has no attribute 'lower'
But sometimes, it returns different errors like:
Traceback (most recent call last): File "/usr/bin/gemBS", line 13, in load_entry_point('gemBS==3.0.0', 'console_scripts', 'gemBS')() File "/usr/lib/python3.4/site-packages/gemBS/commands.py", line 157, in gemBS_main instances[args.command].run(args) File "/usr/lib/python3.4/site-packages/gemBS/production.py", line 112, in run
prepareConfiguration(text_metadata=args.text_metadata,configFile=args.config,no_db=args.no_db,dbfile=args.dbfile,output=args.output) File "/usr/lib/python3.4/site-packages/gemBS/init.py", line 455, in prepareConfiguration db.check() File "/usr/lib/python3.4/site-packages/gemBS/database.py", line 115, in check self.check_contigs(sync) File "/usr/lib/python3.4/site-packages/gemBS/database.py", line 289, in check_contigs for fname, pool, smp, ftype, status in c.execute("SELECT * FROM calling"): ValueError: too many values to unpack (expected 5)
These errors are quite confusing. Are those errors only in older versions and have been fixed now?
For the file structure, it is very simple like this:
-
referenc - Homo_sapiens_assembly38.fa
indexes (5 files generated from "gemBS index" step) - Homo_sapiens_assembly38.contig.sizes / 3. dbSNP_gemBS.idx / gem_indexer_Homo_sapiens_assembly38.BS.gem.err / Homo_sapiens_assembly38.BS.gem / Homo_sapiens_assembly38.BS.info
mapping - IX6859.bam
dbsnp (generated from NCBI ftp server) - *.bed.gz
example.csv
- example.conf
I also attached the configuration files (example.csv and example.conf) I used. I think this is a cool tool and I may make a tiny stupid mistake inside. If you could spend little time and help me have a quick look, I will be grateful a lot.
Thanks, Andy Wang
Research Assistant UBC Heart Lung Innovation | Daley Lab Room 166, 1081 Burrard Street, St. Paul's Hospital, Vancouver, B.C. Canada V6Z 1Y6
test.zip https://github.com/heathsc/gemBS/files/4315344/test.zip
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/heathsc/gemBS/issues/71?email_source=notifications&email_token=AAY4657KMWCUDS2KMBLTL5LRG3KXPA5CNFSM4LFKXBKKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IUB3F7Q, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAY4657DSEMW5ORUQQRMQN3RG3KXPANCNFSM4LFKXBKA .
AndyWangSFU
Hi GemBS team,
Thanks for designing gemBS this cool methylation calling tool. As it claims to perform better than bis-snp, our lab wants to do a snp calling to see how many SNPs we can get (right now, just test for one sample). We have computational nodes with gemBS 3.0.0 and 3.5.0 version.
I tried both versions of gemBS and they both worked well until the index step. However, for the next step "gemBS call" (I already hold the .bam files, so wanna skip the "map" step), they return different errors:
--- For gemBS 3.5.0, it returns:
This makes me a little bit confused. As I checked gemBS's manual guide, the index step will not create a file with .gemBS.ref extension. Anyhow, then I tried to run "gemBS index" again and got this:
I am not sure on what happened, but I guess there may be an error when I update the cluster node, which causes gemBS failing to load related related packages/dependencies. This needs to be double checked. If you know the reason, I'll appreciate a lot.
--- For gemBS 3.0.0, it returns:
But sometimes, it returns different errors like:
These errors are quite confusing. Are those errors only in older versions and have been fixed now?
For the file structure, it is very simple with 4 folders and 2 configuration files like this:
I also attached the configuration files (example.csv and example.conf) I used. I think this is a cool tool and I may make a tiny stupid mistake inside. If you could spend little time and help me have a quick look, I will be grateful a lot. @heathsc @MarcosFernandez
Thanks, Andy Wang
Research Assistant UBC Heart Lung Innovation | Daley Lab Room 166, 1081 Burrard Street | St. Paul's Hospital Vancouver, B.C. Canada, V6Z 1Y6
test.zip