STAR alignment error: ERROR in input reads

Question

[prabin.dm]

Hi,

I am using STAR to align my RNAseq datasets and I am having this error

ReadAlignChunk_processChunks.cpp:115:processChunks EXITING because of FATAL ERROR in input reads: unknown file format: the read ID should start with @ or >

This is my code

module load star/2.5.3a

STAR -- genomeDir mouse/star_genome_mm10 \
        -- readFilesIn L001_R1_001.fastq.gz \
        --outSAMtype BAM SortedByCoordinate \
        --outSAMunmapped Within \
        --twopassMode Basic \
        --outFilterMultimapNmax 1 
        --quantMode TranscriptomeSAM \
        --runThreadN 6 \
        --outFileNamePrefix "STAR_output/Test/"

The Fastq files look like this

@NS500540:133:HNFTLBGX5:1:11101:11802:1042 1:N:0:ACTGAT
CTCCGNTTTATTTATTTGTTCTGCAAATTCGATGCGTCTACCTTCAAATAAAGCATTCATCTTTCTCTGTGACTCT
+
AAAAA#EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE

Is there something wrong I am not able to figure out? I have checked the file for each line, they start with @

thank you

[GenoMax]

I have checked the file for each line, they start with @

Are you certain about that? Did you do anything to this file that may have corrupted the format (e.g. improper trimming)?

You can try validateFiles utility from Jim Kent to see if your file checks out.

[prabin.dm]

Thats looks like an useful utility.

When I downloaded it saves as a text file, which I cant run. Can you please let me know how to use it?

[GenoMax]

You need to add execute permission to it by doing chmod a+x validateFiles before you can run it.

[WouterDeCoster]

There shouldn't be a space in -- readFilesIn. Please select a title which describes your problem better than this.

[prabin.dm]

thank you. I will make that change. Also, I changed the title.

[Pierre Lindenbaum]

what are the outputs of

file L001_R1_001.fastq.gz

and

gunzip -c L001_R1_001.fastq.gz | paste - - - - | cut -c 1 | uniq | sort | uniq

?

[prabin.dm]

the output for file L001_R1_001.fastq.gz is L001_R1_001.fastq.gz: gzip compressed data, extra field

and the output for gunzip -c L001_R1_001.fastq.gz | paste - - - - | cut -c 1 | uniq | sort | uniq is @

Answer 1

[h.mon]

If your fastq files are gzip-compressed, you have to use the parameter --readFilesCommand zcat.

[prabin.dm]

That worked. Thank you

While we are at it, can i ask another question?

I prepared STAR genome indices with --sjdbOverhang 99. Currently the readlength of my fastq files are 75bp.

Should I prepare genome indices again ?

[h.mon]

Biostars is better organized if each thread has only one question. Anyway, no, you do not need to build the indices again, see this post: Confused about sjdbOverhang .

[prabin.dm]

Thanks again. I will do that.