EXITING because of FATAL ERROR in reads input: quality string length is not equal to sequence length

[Laura Spector]

I am running STAR-2.7.1a and I get the error EXITING because of FATAL ERROR in reads input: quality string length is not equal to sequence length. I saw on a previous post the suggestion to use --readMapNumber 1 and, when doing so, my output is shown below:

EXITING because of FATAL ERROR in reads input: quality string length is not equal to sequence length
@M03525:380:000000000-CDJ38:1:1101:13087:1473_GGTGTGCGTGGA
+
@M03525:380:000000000-CDJ38:1:1101:15214:1492_CGTGACATGGGG
SOLUTION: fix your fastq file

It looks like it's entirely missing the quality string and sequence string. The paired end file lengths are the same and divisible by 4. Interestingly, when I run STAR on a copy of the files pre-trimming/barcode extraction (noting that the read IDs are modified slightly upon trimming and barcode extraction by removal of the sample index, i.e., 1:N:0:TCGCCTTA, and addition of the barcode), it works fine.

Here's what I'm running:

~/STAR-2.7.1a/bin/Linux_x86_64$ ./STAR --runMode alignReads --runThreadN 48 --limitBAMsortRAM 10000000000 --readMapNumber 1 --genomeDir STAR_GENOME_overhang100 --readFilesIn DU_1_R1_done.fastq DU_1_R2_done.fastq --alignIntronMax 1 --alignMatesGapMax 2500 --outFileNamePrefix lmpcr_4000/ --outSAMtype BAM SortedByCoordinate > lmpcr_4000/LS501.STARstdout.log

[Alexander Dobin]

Hi Laura,

I suspect there is some kind of formatting problem with fastq files.

could you post (or email me) the first 4 lines of both read1 and read2, i.e.

head -n4 DU_1_R1_done.fastq DU_1_R2_done.fastq

Cheers

Alex

[XX Yang]

Hi Alex,

    I'm running STAR in 2-pass mode for RNA-seq data and error 'EXITING because of FATAL ERROR in reads input: quality string length is not equal to sequence length' was produced. It confuses me that I've ran the STAR for another RNA-seq dataset sequenced by the same company before and it finished successfully.

    The command line I run this time was (STAR version 2.7.1a):

STAR --runMode alignReads --runThreadN 20 --genomeDir /datapool/Index_files/STAR_hg38_index/Homo_sapiens_assembly38.pri  /

--readFilesIn /datapool/home/fck/yangxx/SCZ59/seqdata/RNAseq/SCZ_01.early.R1.fastq.gz /datapool/home/fck/yangxx/SCZ59/seqdata/RNAseq/SCZ_01.early.R2.fastq.gz  /

--readFilesCommand gunzip -c --twopassMode Basic --outFileNamePrefix /datapool/home/fck/yangxx/SCZ59/result/RNAseq/STAR/SCZ_01.early/  /

           --outSAMtype BAM SortedByCoordinate --outSAMattrIHstart 0 --quantMode TranscriptomeSAM

    The first 4 lines of R1 file were extracted by 'less -S /datapool/home/fck/yangxx/SCZ59/seqdata/RNAseq/SCZ_01.early.R1.fastq.gz| head -n4':

@A00808:11:HJKWNDSXX:1:1101:3332:1063 1:N:0:CAAGGAGC+TCTCGCGC

CGCAGCCCACCTCATAAAACCCAGCATTCCCCTCACAGCAGATCACCAGCTTCTGTCCCTGGGGCTCAGCTGTCCCCCGCCGGTCCACAAACATGGTGTCAATCTCATTGCCATCACAGGCCAGCAGCTTTGCCCGGCGCCCATTACACT

+

F:FF,FF:FFFFFFFFFFFFF:FFFFFFFFFFFFFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF,FF:FF

   and first 4 lines of R2 file:

@A00808:11:HJKWNDSXX:1:1101:3332:1063 2:N:0:CAAGGAGC+TCTCGCGC

GGCCAGTCGACTTCCACTGGGAAGAACCCAGCAGCCGGAAGGAGTCTCGAGGGGGCCCTTCCCGCCGGGGTGTGGCCCTGCTTCGCCCAGAGCCCCTGCACCGGGGGACCGCAGACACCCTCCTCAACCGGGTTAAGAAGCTGCCTTGTC

+

FF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF,F,FFFFFF,FFFFFFFFF,FFFFFFFFFFF:FFFFFFFFFFF

    The command line run successfully for another RNA-seq dataset was (STAR version 2.5.2b, did not use 2-pass mode parameters):

STAR --runMode alignReads --runThreadN 20 --genomeDir /home2/yangxx/STAR-index/mm10  /

--readFilesIn /home2/yangxx/visual_cortex/seq_data/Hiseq/BDbi-1_L4_I375.R1.clean.fastq.gz /home2/yangxx/visual_cortex/seq_data/Hiseq/BDbi-1_L4_I375.R2.clean.fastq.gz  /

--outFileNamePrefix /home2/yangxx/visual_cortex/result/STAR/BDbi-1/BDbi-1. --outSAMtype BAM SortedByCoordinate --outSAMattrIHstart 0  /

--readFilesCommand gunzip -c --outSAMstrandField intronMotif --outFilterIntronMotifs RemoveNoncanonical --alignEndsType EndToEnd

      The first 4 lines of BDbi R1 file:

@HISEQ:782:C9Y3HANXX:4:1101:1322:2223 1:N:0:TGAAGCT

TTGCGGTGCACGATGGAGGGGCCGGACTCATCGTACTCCTGCTTGCTGATCCACATCTGCTGGAAGGTGGACAGTGAGGCCAGGATGGAGCCACCGATCCACACAGAGTACTTGCGCTCAGGAGG

+

BBBBBFFFFFFFFFFFFFFFFFFFFFFBFFFFFFFFFFFFFBFFFFFFFFFFFFFFFFFFFFFFFFFFFF<FFFFFBFFFFFFFFFFFFF/<FF<B<</BFFBF<FFFFFFBFF/BBFF/FB/BB

    The first 4 lines of BDbi R2 file: 

@HISEQ:782:C9Y3HANXX:4:1101:1322:2223 2:N:0:TGAAGCT

GAAGTGTGACGTTGACGTCCGTAAAGACCTCTATGCCAACACAGTGCTGTCTGGTGGTACCACCATGTACCCAGGCATTGCTGACAGGATGCAGAAGGAGATTACTGCTCTGGCTCCTAGCACCA

+

BBBBBFFFFFFFFFFF<FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF<FBBBFFF/FFFFFFFFFFFFFFFFFBF/B/BFF<

    Although the read length and sequencing platform for these two datasets were different, I guess it should not influence the running of STAR. What should I do to solve this  problem? 

Best wishes,

Xiaoxue Yang

[Alexander Dobin]

Hi XX Yang,

what was the exact error message from STAR? It lists the read name, sequences and qualities for the specific read that it did not like.

Cheers

Alex

[Jenny McGrady]

Hi, I'm running into this issue as well.

I'm having trouble mapping my reads that were trimmed with Scythe. To troubleshoot, I tried mapping just the first line, but am still getting an error. The error I get with the whole trimmed file (not using the readMapNumber option) is that the read ID does not start with @ or < but the error I get with the single read is that the quality string is not equal to the sequence length. When I look at the trimmed fastq files it doesn't appear that these errors are correct. Also, if I use the same command to map the untrimmed files it works fine. Here is the exact error that I'm getting:

EXITING because of FATAL ERROR in reads input: quality string length is not equal to sequence length

@K00208:YBP029:8:1101:78.29:13.82#0/1
+
@K00208:YBP029:8:1101:88.44:13.82#0/1

SOLUTION: fix your fastq file

Aug 20 15:42:21 ...... FATAL ERROR, exiting

For reference, this is the read I was attempting to map that got the error:

read1

@K00208:YBP029:8:1101:78.29:13.82#0/1
NTCCCGGAGCCGGTCGCGGCGCACCGCCACGGTGGAAGTGCGCCCGGCNGCGNCCGGTNNCNNNCNNGNGNNNNNNNNNNNNNNNNNNNNNNNNCCGGNCC
+
@``eeiii`ieiiiiii`[eiiiiiii`iiiiiiiiii[ie`iiiiei@eii@iiii`@@i@@@i@@i@L@@@@@@@@@@@@@@@@@@@@@@@@V`i[@ee

read2

@K00208:YBP029:8:1101:78.29:13.82#0/3
NAAACCGTTAAGAGGTAAACGGGTGGAGTCCGCGCAGTCCGNCCGGAGGANTCANCACAGCGNNGAGCNANCNNNCNNGNCGNTGNTNCCNGCGGATCTTT
+
@`[[[`e`eeieieLe[iieLei`V`LVeeiii`eieii`V@`iLVLV[i@VeV@LLVLVVV@@VLV[@L@L@@@V@@L@VL@L[@e@eL@LVHL[[`[V`

And here was my command for mapping:
STAR --runMode alignReads --genomeDir ~/project-kmartin/Jenny/mm10_indices/mm10StarIndex/ --runThreadN 64 --readFilesIn ~/project-kmartin/Jenny/RNASeq/demulti/ACSF2_1_trimmed.fastq ~/project-kmartin/Jenny/RNASeq/demulti/ACSF2_2_trimmed.fastq --outFileNamePrefix ~/project-kmartin/Jenny/RNASeq/aligned/ACSF2 --outSAMtype BAM SortedByCoordinate --outWigType bedGraph --genomeLoad LoadAndKeep --limitBAMsortRAM 40000000000 --quantMode GeneCounts --readMapNumber 1

Any help would be greatly appreciated!

--Jenny

[Alexander Dobin]

Hi Jenny,

I think the problem is that the trimmed file is missing the sequence of this read. STAR will not process reads with empty sequences.

What is the output of

grep -A3 ^@K00208:YBP029:8:1101:78.29:13.82  ~/project-kmartin/Jenny/RNASeq/demulti/ACSF2_1_trimmed.fastq

Cheers

Alex