minillinim / mikezbioscripts Goto Github PK

cram files are smaller, maybe 60% or so in my experience for metagenomic data. With access to all the necessary file paths already given, it should be reasonably straightforward to implement.

Only problem is, it might require the user to create a samtools faidx of the subject fasta file first. So now to prepare for a run with multiple sets of mappings, the user has to

• run bwa index
• run samtools faidx
  Perhaps it might be a good idea to implement a "makeSam --prepare" command, that does all of this for the user?

If running asking for cram output, makeSam should check the faidx file exists before running the mapping, so that it doesn't fail after the mapping as this would waste the user's time.

makeSam currently outputs a bunch of stuff on stderr, from bwa (at least). Thus, it is difficult to see when it actually fails, or just some reads fail. This is particularly true when multiple makeSams are being used in parallel (as is common when mapping multiple sets of proteins to a single set of contigs for groopm prep, for instance).

Proposed solution - ask the user to supply a file where the stderr of internal programs like bwa should be redirected to, or default to either piping it to /dev/null or better yet only showing that output when a non-zero return status is detected (ie when a subprocess.check_call fails and throws).

Hey Mike,

I got a problem with this script. Can you please please help? I would like to do a tetranucleotide binning. Thanks!
This was the error message I got -

uqmharoo@keating:~/Phycisphaera/newbler/assembly2-mi93-ml43$ barcodePlotter.py -b out -c 1:4 -l image

barcodePlotter.py
Copyright (C) 2009, 2010 Lauren bragg, Michael Imelfort

This program comes with ABSOLUTELY NO WARRANTY;
This is free software, and you are welcome to redistribute it

under certain conditions: See the source for more details.

***Cluster testing.
you will need to look at the following files:

• heirachical_clusters.pdf
• internal_validation_clusters.pdf
• stability_validation_clusters.pdf

Loading required package: cluster
Loading required package: class

The number of items to be clustered is larger than 'maxitems'
The memory and time required may be excessive, do you wish to continue?
(y to continue, any other character to exit) y
Error in silhouette(cluster, dmatrix = as.matrix(Dist))[, 3] :
incorrect number of dimensions
In addition: Warning message:
In min(interClust, na.rm = TRUE) :
no non-missing arguments to min; returning Inf
Traceback (most recent call last):
File "/srv/whitlam/bio/apps/sw/bioscripts//bin/barcodePlotter.py", line 337, in
clustRTest(opts.numClusts, r)
File "/srv/whitlam/bio/apps/sw/bioscripts//bin/barcodePlotter.py", line 182, in clustRTest
r('intern <- clValid(mymatrix, '+cent_range+', clMethods = c("hierarchical", "kmeans"), validation = "internal");')
File "/srv/whitlam/bio/apps/sw/rpy2/2.2.1/lib/python2.7/site-packages/rpy2/robjects/init.py", line 225, in call
res = self.eval(p)
File "/srv/whitlam/bio/apps/sw/rpy2/2.2.1/lib/python2.7/site-packages/rpy2/robjects/functions.py", line 82, in call
return super(SignatureTranslatedFunction, self).call(_args, *_kwargs)
File "/srv/whitlam/bio/apps/sw/rpy2/2.2.1/lib/python2.7/site-packages/rpy2/robjects/functions.py", line 34, in call
res = super(Function, self).call(_new_args, *_new_kwargs)
rpy2.rinterface.RRuntimeError: Error in silhouette(cluster, dmatrix = as.matrix(Dist))[, 3] :
incorrect number of dimensions