Skip to content

Latest commit

 

History

History
37 lines (30 loc) · 2.78 KB

samplesCsv.md

File metadata and controls

37 lines (30 loc) · 2.78 KB

Sample table

A sample table (e.g. samples.csv) file is used to list the samples included in an analysis run, their library (PCR) index and sample barcode sequences and optional sample-specific analysis parameters.

It is a comma separated file (csv), with a header line (column names), followed by one sample per line. The first column is required to be sample and contains the name of each sample. When starting from a runFolder (bcl)the libIndex column is also required, specifying the sequencing library (PCR index) for this sample (see below). All other columns are optional. Column names are case sensitive!

Column Description Example
sample Sample name Foobar-2
libIndex Illumina library index used with this sample S701P or ATCGGAC
libName Name for the sequencing library / fastq files; default is to use libIndex S701P
barcodes Tn5-plate wells used for this sample 1A-2H
expectedCells Approximate number of single cells in this sample 50000
subsample (Optional) Number of input reads to use for analysis of this sample 100000000
  • sample and libName should consist only of letters, numbers, dash (-) and dot (.)
  • libIndex: Can be one sequence, a ';'-separated list of multiple sequences or a name from references/fastqIndex.tsv.
  • When running from pre-existing fastq file input, libName should match the first part of the fastq file name for this sample, e.g.: Foo1 for Foo1_*.fastq.gz.
  • expectedCells is optional. If it is left out or set to 0, the number will be estimated starting from the number of cell-barcodes with over minUniqueReads (100; see --qcAndArchRParams) reads.
  • subsample is optional. To activate subsampling, provide a target number of reads for the sample here and set --subsample true when running the workflow

Demultiplexing samples within a sequencing library

During analysis the sequencing data is first split into libraries (Fastq files) based on the PCR index. These are given in the libName / libIndex columns. If multiple samples were included in one sequencing library, these are then be demultiplexed based on the tagmentation (Tn5) barcode. In that case the same library is repeated on multiple lines, once for each sample, with the specific tn5-wells for each listed in barcodes. E.g.

sample libIndex barcodes
Foo S701P
Bar-a S703P 1A-2H
Bar-b S703P 3A-3H

The Tn5 wells used for each sample are given in barcodes as either

  • An individual value (1A)
  • A range of wells (1A-2H)
    • Wells are sorted first by number then by letter, i.e. 1A-1H,2A-H,...
    • Note that all ranges are read in column-wise order; e.g. 1A-2C, refers to 1A-1H (all of column 1) plus 2A-2C.
  • A list of values or ranges, separated by semicolon (;) (1A;2A-2D)