Using run_ceres directly

[yesitsjess]

I can't get prepare_ceres_inputs to work consistently with my proxy settings and, on the rare occasion I do, I can't get it to work with my bowtie and samtools:

> prepare_ceres_inputs(inputs_dir="ceres_inputs",
+                      dep_file="ceres_inputs/ceres_LFC_input.gct",
+                      cn_seg_file="ceres_inputs/ceres_CN_input.tsv",
+                      gene_annot_file="example_data/CCDS.current.txt",
+                      rep_map_file="ceres_inputs/ceres_rep_input.tsv",
+                      genome_id="hg19",
+                      chromosomes=paste0("chr", 1:22),
+                      dep_normalize="zmad")
loading dependency data...

Parsed with column specification:
cols(
  Replicate = col_character(),
  CellLine = col_character()
)
loading copy number data...

mapping sgRNAs to the genome...

sh: bowtie: command not found
sh: samtools: command not found
Error in value[[3L]](cond) : 
  failed to open BamFile: file(s) do not exist:
  '/tmp/RtmpfQL4OU/guides.bam'
In addition: Warning messages:
1: In system(bowtie_cmd) : error in running command
2: In system(samtools_cmd) : error in running command

As a result, I've tried to put together the correct data and supply it directly to run_ceres. It fails with error:

> run_ceres(sg_data=sg_data, cn_data=cn_data, 
+           guide_locus=guide_locus, locus_gene=locus_gene, replicate_map=repmap)
Error in dimnames(x) <- dn : 
  length of 'dimnames' [2] not equal to array extent
In addition: There were 50 or more warnings (use warnings() to see the first 50)

and the warnings are:

Warning messages:
1: In mean.default(x, na.rm = T) :
  argument is not numeric or logical: returning NA

Obviously I'm using real data, but I thought dummy data would help you to spot what I'm doing wrong:

# log fold change calc from plasmid of each gRNA in each sample
dum_sg_lfc <- as.matrix(sapply(1:6, function(x) rnorm(4)))
rownames(dum_sg_lfc) <- c("ATCGA", "ATCGT", "ATCGC", "ATCGG")
colnames(dum_sg_lfc) <- c("A1", "A2", "B1", "B2", "C1", "C2")

# log2ratio copy number at each gRNA cut site in each cell line given as chr:pos
dum_cn_lr <- as.matrix(sapply(1:3, function(x) rnorm(4)))
rownames(dum_cn_lr) <- c("1:100", "1:200", "1:300", "1:400")
colnames(dum_cn_lr) <- c("A", "B", "C")

# dummy data using chr:pos as locus, entrez gene id as gene and sample to cell line names
dum_gl <- data.frame(Guide=rownames(dum_sg_lfc), Locus=rownames(dum_cn_lr))
dum_lg <- data.frame(Locus=rownames(dum_cn_lr), Gene=paste0("eg", 1:nrow(dum_cn_lr)))
dum_rep <- data.frame(Replicate=colnames(dum_sg_lfc), CellLine=gsub("[[:digit:]]*", "", colnames(dum_sg_lfc)))

run_ceres(sg_data=dum_sg_lfc, cn_data=dum_cn_lr, 
          guide_locus=dum_gl, locus_gene=dum_lg, replicate_map=dum_rep)

[joshdempster]

Hi Jessica,

It looks like the bowtie and samtools issues you're seeing are just that your shell isn't finding them. The function prepare_inputs accepts arguments for bowtie_exe and samtools_exe. You can just pass the absolute paths to these tools in those arguments.

Best,

Josh

[yesitsjess]

Is there no way to run it directly with inputs as described in my question? I still get errors:

prepare_ceres_inputs(inputs_dir="ceres_inputs",
                     dep_file="ceres_inputs/ceres_LFC_input.gct",
                     cn_seg_file="ceres_inputs/ceres_CN_input.tsv",
                     gene_annot_file="example_data/CCDS.current.txt",
                     rep_map_file="ceres_inputs/ceres_rep_input.tsv",
                     genome_id="hg19",
                     chromosomes=paste0("chr", 1:22),
                     dep_normalize="zmad", 
                     bowtie_exe="/software/bioinf/bowtie2_v2.2.3/bowtie2",
                     samtools_exe="/software/bioinf/samtools_v1.4.1/bin/samtools")

Error: Encountered internal Bowtie 2 exception (#1)
Command: bowtie2_v2.2.3/bowtie2-align-s --wrapper basic-0 -t -p 4 -a -v 0 -f -S hg19 /tmp/RtmpDk2HXp/guides.fa /tmp/RtmpDk2HXp/guides.sam 
(ERR): bowtie2-align exited with value 1
[E::hts_open_format] fail to open file '/tmp/RtmpDk2HXp/guides.sam'
samtools view: failed to open "/tmp/RtmpDk2HXp/guides.sam" for reading: No such file or directory
Error in value[[3L]](cond) : 
  failed to open BamFile: file(s) do not exist:
  '/tmp/RtmpDk2HXp/guides.bam'

[ytakemon]

I'm running into similar issues as well. With the same error message. Any solutions to this?

[ytakemon]

Just an update that I was able to run the example prepare_ceres_inputs()!

In case anyone else stumbles on this issue in the future, I used the following bowtie and samtools versions:

prepare_ceres_inputs(inputs_dir=gecko_inputs_dir,
                     dep_file=gecko_dep_file,
                     cn_seg_file=cn_seg_file,
                     gene_annot_file=gene_annot_file,
                     rep_map_file=gecko_rep_map,
                     chromosomes=paste0("chr", 1:22),
                     dep_normalize="zmad",
                     bowtie_exe="/gsc/software/linux-x86_64-centos7/bowtie-1.2.3/bowtie",
                     samtools_exe="/gsc/software/linux-x86_64-centos7/samtools-1.9/bin/samtools")

[yesitsjess]

Just an update that I was able to run the example prepare_ceres_inputs()!

In case anyone else stumbles on this issue I used the following bowtie and samtools versions:

prepare_ceres_inputs(inputs_dir=gecko_inputs_dir,
                     dep_file=gecko_dep_file,
                     cn_seg_file=cn_seg_file,
                     gene_annot_file=gene_annot_file,
                     rep_map_file=gecko_rep_map,
                     chromosomes=paste0("chr", 1:22),
                     dep_normalize="zmad",
                     bowtie_exe="/gsc/software/linux-x86_64-centos7/bowtie-1.2.3/bowtie",
                     samtools_exe="/gsc/software/linux-x86_64-centos7/samtools-1.9/bin/samtools")

Thanks for sharing this! :)