miwipe / ngslca Goto Github PK

I think we should add examples of all the different types of plots the Rscript produces.

print query sequence in stdout

Make ngsLCA output as Kaiju output file style (https://github.com/bioinformatics-centre/kaiju) that gives you the counts only for the lowest ranks, rather than Kraken2 or Metaphaln that sum up the counts to the highest ranks.

Which is compatible with Phyloseq (https://joey711.github.io/phyloseq/) and Mia (https://microbiome.github.io/OMA/).

ont_otu_table.csv

We still need a output format description of the .lca files, and the Rscript outputted files.

see below parsed seqs which doesn't fit 100% to reference sequence

samtools view CHL_155_12485.sort.bam | grep -e 'HISEQ:50:C6NJJANXX:1:1108:14286:4677' -e 'HISEQ:50:C6NJJANXX:1:1108:14285:32643' -e 'HISEQ:50:C6NJJANXX:1:1108:14287:23673' -e 'HISEQ:50:C6NJJANXX:1:1108:14289:31539' -e 'HISEQ:50:C6NJJANXX:1:1108:14289:54817' | less -S

Only use the maps with lowest editdistance

As of f118505, after having succesfully compiled ngsLCA on Ubuntu 20.04, I get a segmentation fault and empty output files while running sam2lca on of the example file in the bam_files :

See log below:

$ ./ngsLCA -names ncbi_tax_dmp/names.dmp -nodes ncbi_tax_dmp/nodes.dmp -acc2tax ncbi_tax_dmp/nucl_gb.accession2taxid -bam bam_files/SPL_015_1444.fq.plastids.sorted.bam -outnames SPL_015_1444
	-> Will output lca results in file:		'SPL_015_1444.lca'
	-> [thread1] Will read header
	-> Will output lca weight in file:		'SPL_015_1444.wlca'
	-> Will output log info (problems) in file:	'SPL_015_1444.log'
	-> [thread1] Done reading header: 0.00 sec, header contains: 4322
Segmentation fault (core dumped)

Hi,

Is it possible to combine each taxa rank otutable into one rather than having them separate (one for species, one for genera etc)?

I can see that the program creates files containing all the ranks put together (kraken style), like "complete profile" file or the one in the taxa_groups folder but the counts, in my case, do not match with the counts in the separate files, therefore I am confused.

Thanks

Make ngsLCA output as Kaiju output file style (https://github.com/bioinformatics-centre/kaiju) that gives you the counts only for the lowest ranks, rather than Kraken2 or Metaphaln that sum up the counts to the highest ranks.

Which is compatible with Phyloseq (https://joey711.github.io/phyloseq/) and Mia (https://microbiome.github.io/OMA/).

ont_otu_table.csv

Hi,
I'm wondering if it's ok to run ngsLCA with the bam file from bwa-mem. I see in the tutorial using bowtie2 with some specific parameters. If we use bwa-mem, is there any option that are recommended.
Thanks,
Hien

I've been having problems running ngsLCA on my own data so I tried running the test data first, but I also can't get that to work. I downloaded SPL_015_1444.fq.plastids.bam from ERDA, downloaded the ncbi tax_tax_dmp and then ran the following code:

ngsLCA -editdistmin 0 -editdistmax 0 -names ncbi_tax_dmp/names.dmp -nodes ncbi_tax_dmp/nodes.dmp -acc2tax ncbi_tax_dmp/nucl_gb.accession2taxid.gz -bam SPL_015_1444.fq.plastids.bam -outnames outfile.ed0

filenames needs to be possible to specify

Hi @miwipe ,
While working ont the review of ngsLCA, I tried to compile it following the README's instruction (as of f118505), however, I encountered some errors (see log attached).
Compilation worked fine on Ubuntu 20.04

ngsLCA_build_log.txt

hi:
thanks for ngsLCA
but i get a error when Install the package "ngsLCA" with the command devtools::install_github("wyc661217/ngsLCA").
ERROR: dependencies ‘ComplexHeatmap’, ‘ggpubr’, ‘vegan’ are not available for package ‘ngsLCA’,my R version is latest 4.2.0

Can we add accession number and query sequence length to output print?