Hoohm / dropSeqPipe

A SingleCell RNASeq pre-processing snakemake workflow
Creative Commons Attribution Share Alike 4.0 International
147 stars 47 forks source link

Generate Plot | Error in mmm #6

Closed abmmki closed 7 years ago

abmmki commented 7 years ago

Hi,

In generate-plot step I am getting following error. What I am mising? Using 023a version.

Plotting knee plots Warning messages: 1: Removed 716167 rows containing missing values (geom_point). 2: Removed 563180 rows containing missing values (geom_point). 3: Removed 1336568 rows containing missing values (geom_point). 4: Removed 1230829 rows containing missing values (geom_point). 5: Removed 1180851 rows containing missing values (geom_point). 6: Removed 1473024 rows containing missing values (geom_point). Plotting base stats Loading required package: magrittr Error in mmm < each : comparison of these types is not implemented Calls: plotRNAMetrics ... Reduce -> f -> rbind_gtable -> compare_unit -> unit -> comp Execution halted Traceback (most recent call last): File "/bin/dropSeqPipe", line 9, in load_entry_point('dropSeqPipe==0.23a0', 'console_scripts', 'dropSeqPipe')() File "/usr/lib/python3.4/site-packages/dropSeqPipe/main.py", line 182, in main shell(base_summary) File "/usr/lib/python3.4/site-packages/snakemake-3.10.1-py3.4.egg/snakemake/shell.py", line 80, in new raise sp.CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command 'Rscript /usr/lib/python3.4/site-packages/dropSeqPipe/Rscripts/singleCell/rna_metrics.R /PROJECTS/' returned non-zero exit status 1

abmmki commented 7 years ago

just to add, I am using drosophila BDGP 6 version genome (with many new and uncharacterised chromosomes)

duyck commented 7 years ago

Here's a fix I created. This forces the gtables objects to take the "last" plots width as the width for both.

Edit file: ${HOME}/.local/lib/python3.5/site-packages/dropSeqPipe/Rscripts/singleCell/rna_metrics.R

Replace: grid::grid.draw(rbind(gp1, gp2))
with
grid::grid.draw(rbind(gp1, gp2, size = "last"))

abmmki commented 7 years ago

@duyck Hi, Many thanks. It worked perfectly. Thanks once again.