DEseq2 input files

Question

[sophialovechan]

Hi, I am using HIAST2, Stringtie and DEseq to do differential expression analysis. When I runned DEseq2, I got an error like that Error in DESeqDataSet(se, design = design, ignoreRank) : some values in assay are negative

So I checked my input (counting files), it shows

         X                 sample1  sample2  sample3   sample4 
[97291,] "ENSG00000235759" "     0" "     0" "      0" "     0"
[97292,] "ENSG00000235754" "     0" "     0" "      0" "     0"
[97293,] "ENSG00000235756" "     0" "     0" "      9" "     2"
[97294,] "ENSG00000235757" "     0" "     0" "      0" "     0"
[97295,] "MSTRG.19084"     "     3" "     3" "     12" "    11"
[97296,] "MSTRG.32274"     "     0" "     0" "      0" "     0"

It seems there are a space behind every numbers. I don't know why they are there and how can I get rid of them? Thanks,

[WouterDeCoster]

How did you generate the table you are showing us? Add all commands.

[sophialovechan]

I used prepDE.py to generate the reads counting data.

[sophialovechan]

1. HISAT2--alignment

   module load hisat2/2.1.0
   time hisat2 -p 8 -q --phred33 -x /HISAT2_index/
   -1 SRR4011898_1.fastq -2 SRR4011898_2.fastq \
   -S H9_day0_rep1.sam
   

2. Stringtie--assemble

   stringtie ./H9_day0_rep1.bam \
   -G /Bowtie2Index/genome.gff \
   -p 4 --fr -o H9_Day0_rep1.gtf -l H9_Day0_rep1
   

3. Stringtie---merge

   module load stringtie/1.3.3b
   stringtie --merge -p 8 -G  genes.gtf -o H9_Day0_Day4_merged.gtf mergelist.txt
   

4. generate ballgown

   B -p 8 -G H9_Day0_Day4_merged.gtf -o ballgown/H9_Day0_rep1/H9_Day0_rep1.gtf ./H9_0_rep1/accepted_hits.bam
   

5. prepDE.py

   module load stringtie/1.3.3b
   prepDE.py
   

[Devon Ryan]

I've done what I could to reformat your post so the command render in a meaningful way. I also fixed what I presume was a type in step 2 (an extra ). There's one remaining issue in step 4 (ballgown) where it looks like your full command didn't get pasted in. I don't know what your full command was, so I can't fix that.

Answer 1

[Devon Ryan]

At least for me, read.delim() seems to be smart enough to strip the white space surrounding the numbers. But anyway a simple sed -ei "s/ //g" input_file.txt will remove the extra spaces (I assume there are tabs in the original file as well).