ONT Tools can analyse multiple samples in a single execution. All samples will be analysed with the same configuration.
This pipeline is designed to run on a raw data folder of any ONT sequencing platform. The data can be available locally or, via a mounted network drive, still be on the machine. Before analysis, the reads are copied into the project folder. The raw data folders are not modified in any way.
ONT Tools normally requires a File of Filenames (FoF) to associate sample IDs and raw data locations. By default, the pipeline will look for a file named sample_run_table.tsv in the working directory (can be changed in options file). The FoF must be a Tab separated text file with two or three columns:
- Sample ID
- Data folder
- (Optional) Barcode
The sample ID can be defined by the user and will be used for output files and reports. The data folder should be in ONT run folder format and contain subfolders for fail and pass reads. The sequencing summary file must be located directly in this folder. Other folder structures, for an example a raw folder containing only FASTQs should work but might be buggy. The barcode column must contain the same names as the barcode subdirs in the run folder.
A single sample can have multiple data folders, this is used when multiple flowcells were used for the same sample, or if the sequencer created multiple runfolders because of interruptions. In this case there are multiple rows with the sampleID.
| SampleID | Folder |
|---|---|
| a0001 | /mnt/ontdata/run_20201205 |
| a0001 | /mnt/ontdata/run_20201205_restarted |
| a0002 | /mnt/ontdata/run_20201206 |
Multiple samples can be sequenced in the same run using barcodes. Normally, the raw_data folder already containes subfolders for the different barcodes. In this case, the FOFN should contain folder locations to the correct barcode subdirectories.
| SampleID | Folder | Barcode |
|---|---|---|
| a0001 | /mnt/ontdata/run_20201205 | barcode01 |
| a0001 | /mnt/ontdata/run_20201205_restarted | barcode01 |
| a0002 | /mnt/ontdata/run_20201205 | barcode02 |
A separate PycoQC will be created for each row in the Filelist.
The paths in the Folder column can be absolute or relative to the working directory. They will be searched recursively for all files ending with .fastq, .fastq.gz, .fq , .fq.gz and combined into a single compressed FastQ file in the sample directory.
By default, subfolders with the name failed will be excluded from the search. This behaviour can be changed by setting use_failed_reads = True in the config.yaml file.
Of there is no valid FoF input, ONT Tools tries to gather Sample_IDs from existing folders. In this case, every Folder with a Sample_ suffix will be searched for possible ONT analyses.