Chunks erros with STAR aligner

Question

[Morris_Chair]

Hello everyone, I’m always “lucky” to catch all the kind of errors in my paths… I run star as I did in the past without many problems, but with this files I have a new issue.. it might be because the reads are too long? Does anybody dealt with this problem before?

STAR --genomeDir genome/Indexed_star/ --runThreadN 4 --readFilesIn /home/RNAseq/fastq_files/file.R1.fq --outFileNamePrefix mapping/star/file.R1. --outSAMattributes All --outSAMtype BAM SortedByCoordinate --quantMode GeneCounts --outFilterMismatchNmax 100



Apr 09 20:17:44 ..... started STAR run
Apr 09 20:17:44 ..... loading genome
Apr 09 20:19:42 ..... started mapping

ReadAlignChunk_processChunks.cpp:171:processChunks EXITING because of FATAL ERROR in input reads: unknown file format: the read ID should start with @ or >

Apr 09 20:19:43 ...... FATAL ERROR, exiting
Segmentation fault (core dumped)

The read start with @ anyways, after trimming I woulnd't expect does N ..

@K00124:54:HTMMCBBXX:1:1101:8968:993 1:N:0:NCGTCC
NGCATCTCAATTTGGAATCCATTAATTCATCAGGTTTTGCCTCTTTCCACACAGCTTCCATATCTGAAGTGTTTGGTGGAGCAAAAATA
+

Thank you for help

[Pierre Lindenbaum]

what are the outputs of

file /home/RNAseq/fastq_files/file.R1.fq

cat /home/RNAseq/fastq_files/file.R1.fq | paste - - - - | wc -l

cat /home/RNAseq/fastq_files/file.R1.fq | paste - - - - | cut -f 1| cut -c 1 | uniq | sort | uniq

[Morris_Chair]

Hi Pierre, This are:

cat file.R1.fq | paste - - - - | wc -l
22913861

file file.R1.fq
file.R1.fq: ASCII text

cat file.R1.fq | paste - - - - | cut -f 1| cut -c 1 | uniq | sort | uniq
@

Merci

[Pierre Lindenbaum]

and wc -l file.R1.fq please

[Morris_Chair]

Hi Pierre, I also counted the number of reads and the number of @ to see if there are reads with no @ and the number is the same

echo $(cat file.R1.fq|wc -l)/4|bc
30816907
grep "@" file.R1.fq | wc -l
30816907

[GenoMax]

The read start with @ anyways, after trimming I woulnd't expect does N ..

Not true. Unless you had trimmed to remove N's they remain. Some reads at the beginning of the file may have these N's (or are they present in each and every read)?

[Morris_Chair]

No they are not present in all the reads but for sure on the first 100 reads or so

thanks

[Morris_Chair]

wc -l file.R1.fq
91655444 file.R1.fq

tnx

Answer 1

[Morris_Chair]

Problem Solved !!! I installed the most updated version of STAR,

STAR-2.7.0f.tar

I hope this will help other people too

Cheers,