FASTQ header mismatch detected

[StackChan]

6
SRR21422687
90
4.50 G
1.17 Gb
SRX17427003
CRC_16_S1_L001_R2_001
Reads are like follows:
>gnl|SRA|SRR21422687.1.1VH00699:3:AAAHK77HV:1:2412:56765:48347 Biological (Biological)
AAGCAGTGGTATCAACGCAGAGTACATGGGGACAGACCCGGAGAGCACCGCGAGGGCGGAGCTGCGTTCTCCTCTGCACAGATTTCGGTG

8
SRR21422689
28
1.40 G
598.12 Mb
SRX17427001
CRC_16_S1_L001_R1_001
Reads are like follows:
>gnl|SRA|SRR21422689.1.1VH00699:3:AAAHK77HV:1:2412:56765:48347 Biological (Biological)
ATGCCATTTGCGACCAGAGAGGTGAATT

So when you run cellranger or spaceranger, you will encounter "FASTQ header mismatch detected" problems because their header is different and shall both be like .

>gnl|SRA|SRR21422687.1.1VH00699:3:AAAHK77HV:1:2412:56765:48347 Biological (Biological)

It really sucks and I guess it's because the author split a big SRA like CRC_16's SRA to 8 SRAs and their fastaq headers are renamed.
Not so many people really reproduce the paper and report their issues here.
The same problem is encountered by issue #15 .

[nicolas-zimmermann]

I encounter the same problem:

2024-12-03 17:39:43 [runtime] (chunks_complete) ID.crc16.SPATIAL_RNA_COUNTER_CS.SPATIAL_RNA_COUNTER.DISABLE_FEATURE_STAGES
2024-12-03 17:39:43 [runtime] (failed)          ID.crc16.SPATIAL_RNA_COUNTER_CS.SPATIAL_RNA_COUNTER_PREP.DETECT_CHEMISTRY

[error] Pipestance failed. Error log at:
crc16/SPATIAL_RNA_COUNTER_CS/SPATIAL_RNA_COUNTER_PREP/DETECT_CHEMISTRY/fork0/chnk0-uaac34f344f/_errors

Log message:
FASTQ header mismatch detected at line 4 of input files "crc/fastq/CRC_16_S1_L001_R1_001.fastq.gz" and "crc/fastq/CRC_16_S1_L001_R2_001.fastq.gz": file: "crc/fastq/CRC_16_S1_L001_R1_001.fastq.gz", line: 4

Waiting 6 seconds for UI to do final refresh.
Pipestance failed. Use --noexit option to keep UI running after failure.

2024-12-03 17:39:49 Shutting down.

Or perhaps fastq-dump or fasterq-dump should be perform otherwise on the SRA, or another step should be done before processing the fastq.
Also I renamed the fastq manually in order to run spaceranger.

Any help would be greatly appreciated !

Thanks in advance.
Best

[nicolas-zimmermann]

Also counting reads in each file with awk indicate different count numbers for R1 and R2 of the same samples :

$ awk 'END {print NR/4}' CRC_16_S1_L001_R1_001.fastq.gz
1.07459e+06

$ awk 'END {print NR/4}' CRC_16_S1_L001_R2_001.fastq.gz
1.4658e+06 

Maybe I'm doing it wrong and there is a way around this issue ?

[StackChan]

FASTQ provided by the author has header mismatch detected indeed.

So you can use sed to uniform their header, like as follows,
(It will definitely work for you cauz the reason of this error is clear and easy to fix. It works for me and I have got the expected result.)

#!/bin/bash

declare -A header_map_r1
header_map_r1=(
    ["CRC_S1_L001_R1_001"]="SRR21422689"
    ["CRC_S1_L001_R1_002"]="SRR21422688"
    ["CRC_S1_L002_R1_001"]="SRR21422685"
    ["CRC_S1_L002_R1_002"]="SRR21422684"
)

declare -A header_map
header_map=(
    ["CRC_S1_L001_R1_001"]="SRR21422687"
    ["CRC_S1_L001_R1_002"]="SRR21422686"
    ["CRC_S1_L002_R1_001"]="SRR21422683"
    ["CRC_S1_L002_R1_002"]="SRR21422682"
)

input_dir="/data/pub/SRT/CRC_16_10x_Visium_ST/raw/"
output_dir="${input_dir}/modified/"

mkdir -p "$output_dir"

for r1_file in "${!header_map[@]}"; do
    r1_srr="${header_map_r1[$r1_file]}"
    r2_srr="${header_map[$r1_file]}"
    input_file="${input_dir}${r1_file}.fastq.gz"
    output_file="${output_dir}${r1_file}.fastq.gz"

    if [ -f "$input_file" ]; then
        echo "Processing $input_file..."
        
        zcat "$input_file" | sed "s/^@${r1_srr}/@${r2_srr}/" | gzip > "$output_file"
        
        echo "Modified file saved to $output_file."
    else
        echo "Input file $input_file not found!"
    fi
done