fastF aims to sample, filter, and summarise single-cell RNA sequencing result, FastQ file. It consists 5 subcommands for the specific tasks.

> ./fastF --help
Usage: fastF <subcommands> [args]

subcommands:

       freq:    Find all the cell barcode whitelist and their frequencies.
       filter:  Filter fastq file using cell barcode whitelist and read depth.
       crb:     Extract CR and CB tags from bam file and summarize them with 
                    frequencies to a tsv file.
       bam2db:  Filter bam file with desired cell proportion and read depth, 
                    then summarise it UMI matrix.
       extract: Extract the tag of bam file and write it to tag_summary.csv.


   -h, --help

freq: Find all the cell barcode whitelist and their frequencies.

• Compare to umi-tools whitelist, it extracts all the unique cell barcodes and their frequencies.

> ./fastF freq --help

Find all the cell barcode whitelist and their frequencies.

    -h, --help        

Basic options
    -R, --R1=<str>    path to R1 fastq files
    -o, --out=<str>   path to output whitelist
    -l, --len=<int>   length of cell barcode
    -u, --umi=<int>   length of UMI

filter: Filter fastq file using cell barcode whitelist and read depth.

• It is the motivation of the entire project, see the similar questions arised in the community, and FastQDesign.

• When filter through the assumed read depth, it is compare to seqtk, but it could process all I1, R1 and R2 concurrently as needed.

• Even more, it could filter FastQ reads according to the given cell barcode whitelist.

• Of course, whitelist and read depth can be set at the same time.

> ./fastF filter --help
Filter fastq file using cell barcode whitelist and read depth.

    -h, --help                

Basic options
    -I, --I1=<str>            optional, path to sample I1 fastq files
    -R, --R1=<str>            required, path to sample R1 fastq files
    -r, --R2=<str>            optional, path to sample R2 fastq files
    -o, --out=<str>           dir to output fastq files
    -w, --whitelist=<str>     whitelist of cell barcodes
    -l, --len=<int>           length of cell barcode
    -s, --seed=<int>          seed for random number generator
    -t, --rate=<flt>          rate of reads to keep after matching cell barcodes
    -a, --allcells            keep all reads with cell barcode

crb: Extract CR and CB tags from bam file and summarize them with frequencies to a tsv file.

• Once again, this command is very compared to umi-tools whitelist. Instead of calculating the distance between all the raw cell barcodes, and clustering, it directly uses the tags CR and CB information in the bam file produced by cellranger.

• The result is summarise by the corrected cell barcode with its variants, along with their frequencies.

> ./fastF crb --help

Extract CR and CB tags from bam file and summarize them with frequencies to a tsv file.

    -h, --help        
    -b, --bam=<str>   path to bam file
    -o, --out=<str>   path to output directory

bam2db: Filter bam file with desired cell proportion and read depth, then summarise it into UMI matrix.

• bam2db is an upgraded version of filter, it can automatically find whitelist and subsample cell with desired read depth.

• Compared to filter, it won't produce the filtered raw FastQ read, instead, it summarises the filtered result as UMI matrix.

• Not to mention, it is extremely computionaly efficient.

> ./fastF bam2db --help

Filter bam file with desired cell proportion and read depth with sqlite3, then summarise it UMI matrix.

    -h, --help            
    -b, --bam=<str>       path to bam file
    -f, --feature=<str>   path to feature list file
    -a, --barcode=<str>   path to barcode list file
    -d, --dbname=<str>    name of database
    -c, --cell=<flt>      rate of cell barcode
    -r, --depth=<flt>     rate of depth
    -o, --out=<str>       path to output directory
    -s, --seed=<int>      seed for random number generator

extract: Extract the tag of bam file and write it to tag_summary.csv.

• This command just can help extract any tag from the bam file.

> ./fastF extract --help

Extract the tag of bam file.

    -h, --help        
    -b, --bam=<str>   path to bam file
    -t, --tag=<str>   tag of bam file
    -T, --type=<int>  type of tag, 0: string, 1: integer