bcl2fastq > cbcl files not found

[plhm]

Hello there all.

Snakemake is trying to run the bcl2fastq command to obtain fastqs from the bcl files, but I got the following error when doing so:

2023-12-15 14:03:10 [2b8569c9a180] INFO: Patterned flowcell detected
2023-12-15 14:03:10 [2b8569c9a180] WARNING: No cbcl files found for cycle 1. Attempt 1 of 3. Retry with cycle 2.
2023-12-15 14:03:10 [2b8569c9a180] WARNING: No cbcl files found for cycle 2. Attempt 2 of 3. Retry with cycle 3.
2023-12-15 14:03:10 [2b8569c9a180] ERROR: bcl2fastq::common::Exception: 2023-Dec-15 14:03:10: Success (0): /opt/conda/conda-bld/bcl2fastq2_1548424849859/work/src/cxx/lib/layout/Layout.cpp(1028): Throw in function static void bcl2fastq::layout::TileLayoutDetector::readTilesFromCbclHeaders(const boost::filesystem::path&, const bcl2fastq::config::SampleSheetCsv&, const std::vector<boost::basic_regex<char, boost::regex_traits<char> > >&, const string&, std::vector<bcl2fastq::layout::LaneInfo>&, bcl2fastq::common::TileFileMap&, bcl2fastq::common::NumBasesPerByte&, bcl2fastq::common::CycleNumber, bcl2fastq::common::CycleNumber, bool)
Dynamic exception type: boost::exception_detail::clone_impl<bcl2fastq::common::InputDataError>
std::exception::what:
No cbcl files found for cycle 3. Attempt 3 of 3.
Corrupt or missing .cbcl files were found for 3 consecutive cycles, verify the source of the data.

Here is the function I ran:

bcl2fastq --barcode-mismatches 1 --ignore-missing-positions --ignore-missing-controls --ignore-missing-filter --ignore-missing-bcls --no-lane-splitting --minimum-trimmed-read-length 15 --mask-short-adapter-reads 15 -R /project/kyathit/RawData/231208_sciseq-Pilot/usftp21.novogene.com/20231128_lh00328_0116_A22GHJ3LT3-L3 --sample-sheet rerio_sci_test/raw_reads/fake.csv --output-dir rerio_sci_test/raw_reads/ --loading-threads 1 --processing-threads 1 --writing-threads 1

And here is how the fake.csv looked:

[DATA] Lane,Sample_ID,Sample_Name,index,index2 ,fake,fake,NNNNNNNNNN,NNNNNNNNNN

I am wondering if this is an issue you've ran into, or whether the bcl file I received from the company is indeed corrupt.

Best,

Pietro

[J0bbie]

Hi Pietro,

Hmm, never seen that before. Is it missing some files or folders? Just to give you an idea of what a typical structure looks like for our data:

I'm also adding the start from .fq.gz this week, that might already alleviate this for you.

[plhm]

Hey, Jobbie.

I figured out what was up. In short, here is how you make your dummy file for bcl2fastq:

# Generate fake sample-sheet to allow indexes to be added to R1/R2. mkdir -p {params.path_out} echo -e "[DATA]\nLane,Sample_ID,Sample_Name,index,index2\n,fake,fake,NNNNNNNNNN,NNNNNNNNNN" > {output.sample_sheet} """

I dug around the bcl2fastq file and found that in not giving the program a lane (i.e. \n,fake), it tries to run bcl2fastq across all lanes in the presumed run. The issue is that I only used one lane out of the idk how many lanes in the flow cell form Novogene.

I fixed the issue by going into your step1_bcl2fastq.smk file, and hardcoding 003 in your echo line (turns out that my lane, was lane # 3). That fixed the issue.

As you said, your fastq option will likely fix this, but, if you have others having the same issue, then may I suggest that you either get a line of code to check which lanes are in the bcl folder, or you ask the user to tell the program what lanes are in the dataset.