4. The survey module¶

The survey module has two main functions. 1) Preprocessing of the library data and 2) assessment of the assemble-ability of each sample/library.

In detail, the steps performed by the survey module are:

1. Preprocessing with bbduk

   Each paired-end library is preprocessed with bbduk with the following operations:

   • k-mer based adapter trimming against bbduk’s default adapter list
   • gentle quality trimming on either side, keeping only bases with at least phred=3
   • length filtering (100bp)
   • quality filtering (maq=20)
   • tpe/tbo (s. bbduk documentation)

   These operations can be adjusted via the bgrrl_config.yaml.

2. Normalization with bbnorm (optional, switch off with --no-normalization)

   Libraries are normalized to the range of 2x-100x.

   The range can be adjusted via the bgrrl_config.yaml.

3. Read quality assessment with fastqc

4. Read sequence feature assessment with kat

5. Survey assembly with tadpole

   Each library is subjected to a simple and quick assembly with tadpole. This determines if the library can be assembled or not. Library assembly statistics are then calculated with quast via qaa.

6. Filtering by assemble-ability

   Libraries that could be assembled with tadpole and that contain at least 1000 reads and have an assembly size above 1Mbp (or user-specified, s. below) will be automatically passed on to the assembly stage. Libraries that fail these checks will be filtered out. However, the user can manually add them to the assembly samplesheet (s. below).

7. Assembly samplesheet generation

   Samples that passed the filtering stage will be written to a new samplesheet (samplesheet.qc_pass.csv) in the <outdir>/reports/samplesheets directory. This samplesheet can then be used as input for the assembly stage.

8. Read packaging

   Preprocessed reads will be automatically packaged into the <outdir>/Data_Package directory unless suppressed by the --no-packaging option.