scruff: Run scruff pipeline

Description Usage Arguments Value Examples

View source: R/scruff.R

Description

Run the scruff pipeline. This function performs all demultiplex, alignRsubread, and countUMI functions. Write demultiplex statistics, alignment statistics, and UMI filtered count matrix in output directories. Return a SingleCellExperiment object containing the count matrix, cell and gene annotations, and all QC metrics.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
scruff(project = paste0("project_", Sys.Date()), experiment, lane, read1Path,
  read2Path, bc, index, reference, bcStart, bcStop, bcEdit = 0, umiStart,
  umiStop, umiEdit = 0, keep, cellPerWell = 1, unique = FALSE,
  nBestLocations = 1, minQual = 10, yieldReads = 1e+06,
  alignmentFileFormat = "BAM", demultiplexOutDir = "./Demultiplex",
  alignmentOutDir = "./Alignment", countUmiOutDir = "./Count",
  demultiplexSummaryPrefix = "demultiplex",
  alignmentSummaryPrefix = "alignment", countPrefix = "countUMI",
  logfilePrefix = format(Sys.time(), "%Y%m%d_%H%M%S"),
  overwrite = FALSE, verbose = FALSE, cores = max(1,
  parallel::detectCores() - 2), threads = 1, ...)

Arguments

project

The project name. Default is paste0("project_", Sys.Date()).

experiment

A character vector of experiment names. Represents the group label for each FASTQ file, e.g. "patient1, patient2, ...". The number of cells in a experiment equals the length of cell barcodes bc. The length of experiment equals the number of FASTQ files to be processed.

lane

A character or character vector of flow cell lane numbers. If FASTQ files from multiple lanes are concatenated, any placeholder would be sufficient, e.g. "L001".

read1Path

A character vector of file paths to the read1 FASTQ files. These are the read files with UMI and cell barcode information.

read2Path

A character vector of file paths to the read2 FASTQ files. These read files contain genomic sequences.

bc

A vector of pre-determined cell barcodes. For example, see ?barcodeExample.

index

Path to the Rsubread index of the reference genome. For generation of Rsubread indices, please refer to buildindex function in Rsubread package.

reference

Path to the reference GTF file. The TxDb object of the GTF file will be generated and saved in the current working directory with ".sqlite" suffix.

bcStart

Integer or vector of integers containing the cell barcode start positions (inclusive, one-based numbering).

bcStop

Integer or vector of integers containing the cell barcode stop positions (inclusive, one-based numbering).

bcEdit

Maximally allowed Hamming distance for barcode correction. Barcodes with mismatches equal or fewer than this will be assigned a corrected barcode if the inferred barcode matches uniquely in the provided predetermined barcode list. Default is 0, meaning no cell barcode correction is performed.

umiStart

Integer or vector of integers containing the start positions (inclusive, one-based numbering) of UMI sequences.

umiStop

Integer or vector of integers containing the stop positions (inclusive, one-based numbering) of UMI sequences.

umiEdit

Maximally allowed Hamming distance for UMI correction. For read alignments in each gene, by comparing to a more abundant UMI with more reads, UMIs having fewer reads and with mismatches equal or fewer than umiEdit will be assigned a corrected UMI (the UMI with more reads). Default is 0, meaning no UMI correction is performed. Doing UMI correction will decrease the number of transcripts per gene.

keep

Read trimming. Read length or number of nucleotides to keep for read 2 (the read that contains transcript sequence information). Longer reads will be clipped at 3' end. Shorter reads will not be affected. This number should be determined based on the sequencing kit that was used in library preparation step.

cellPerWell

Number of cells per well. Can be an integer (e.g. 1) indicating the number of cells in each well or an vector with length equal to the total number of cells in the input alignment files specifying the number of cells in each file. Default is 1.

unique

Argument passed to align function in Rsubread package. Boolean indicating if only uniquely mapped reads should be reported. A uniquely mapped read has one single mapping location that has less mis-matched bases than any other candidate locations. If set to FALSE, multi-mapping reads will be reported in addition to uniquely mapped reads. Number of alignments reported for each multi-mapping read is determined by the nBestLocations parameter. Default is FALSE.

nBestLocations

Argument passed to align function in Rsubread package. Numeric value specifying the maximal number of equally-best mapping locations that will be reported for a multi-mapping read. 1 by default. The allowed value is between 1 to 16 (inclusive). In the mapping output, "NH" tag is used to indicate how many alignments are reported for the read and "HI" tag is used for numbering the alignments reported for the same read. This argument is only applicable when unique option is set to FALSE.

minQual

Minimally acceptable Phred quality score for cell barcode and UMI sequences. Phread quality scores are calculated for each nucleotide in these tags. Tags with at least one nucleotide with score lower than this will be filtered out. Default is 10.

yieldReads

The number of reads to yield when drawing successive subsets from a fastq file, providing the number of successive records to be returned on each yield. This parameter is passed to the n argument of the FastqStreamer function in ShortRead package. Default is 1e06.

alignmentFileFormat

File format of sequence alignment results. "BAM" or "SAM". Default is "BAM".

demultiplexOutDir

Output folder path for demultiplex results. Demultiplexed cell specifc FASTQ files will be stored in folders in this path, respectively. Make sure the folder is empty. Default is "./Demultiplex".

alignmentOutDir

Output directory for alignment results. Sequence alignment maps will be stored in folders in this directory, respectively. Make sure the folder is empty. Default is "./Alignment".

countUmiOutDir

Output directory for UMI counting results. UMI filtered count matrix will be stored in this directory. Default is "./Count".

demultiplexSummaryPrefix

Prefix for demultiplex summary filename. Default is "demultiplex".

alignmentSummaryPrefix

Prefix for alignment summary filename. Default is "alignment".

countPrefix

Prefix for UMI filtered count matrix filename. Default is "countUMI".

logfilePrefix

Prefix for log file. Default is current date and time in the format of format(Sys.time(), "%Y%m%d_%H%M%S").

overwrite

Boolean indicating whether to overwrite the output directory. Default is FALSE.

verbose

Boolean indicating whether to print log messages. Useful for debugging. Default to FALSE.

cores

Number of cores to use for parallelization. Default is max(1, parallel::detectCores() - 2), i.e. the number of available cores minus 2.

threads

Do not change. Number of threads/CPUs used for mapping for each core. Refer to align function in Rsubread for details. Default is 1. It should not be changed in most cases.

...

Additional arguments passed to the align function in Rsubread package.

Value

A SingleCellExperiment object.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
## Not run: 
# prepare required files

data(barcodeExample, package = "scruff")
fastqs <- list.files(system.file("extdata", package = "scruff"),
    pattern = "\\.fastq\\.gz", full.names = TRUE)
fasta <- system.file("extdata", "GRCm38_MT.fa", package = "scruff")
gtf <- system.file("extdata", "GRCm38_MT.gtf", package = "scruff")

# NOTE: Rsubread package does not support Windows environment.
library(Rsubread)
# Specify the basename for Rsubread index
indexBase <- "GRCm38_MT"
# Create index files for GRCm38_MT.
buildindex(basename = indexBase, reference = fasta, indexSplit = FALSE)

# run scruff pipeline
sce <- scruff(project = "example",
    experiment = c("1h1", "b1"),
    lane = c("L001", "L001"),
    read1Path = c(fastqs[1], fastqs[3]),
    read2Path = c(fastqs[2], fastqs[4]),
    bc = barcodeExample,
    index = indexBase,
    reference = gtf,
    bcStart = 1,
    bcStop = 8,
    umiStart = 9,
    umiStop = 12,
    keep = 75,
    cellPerWell = c(rep(1, 94), 0, 0, rep(1, 94), 300, 1),
    overwrite = TRUE,
    verbose = TRUE)

## End(Not run)

# or use the built-in SingleCellExperiment object generated using
# example dataset (see ?sceExample)
data(sceExample, package = "scruff")

compbiomed/scruff documentation built on May 30, 2019, 12:48 p.m.