What is pears?

Installing pears

1. git clone https://github.com/DavidsonGroup/pears.git
2. make sure you have nextflow installed. See docs.
3. Specify your parameters in the nextflow.config file
4. run nextflow run pears -c /path/to/nextflow.config

Input

Pears requires:

• single-cell paired-end reads
• known list of fusions
• reference genome

input	requirements
Reads	named according to cellranger: [Sample Name]_S1_L00[Lane Number]_[Read Type]_001.fastq.gz
Cell barcode and Unique Molecular Barcode (UMI)	You will need to specify the length of the cell barcode and UMI in your data

Your known list of fusions should contain:

• name of gene fusion
• for both genes: chromosome, base location of breakpoint, strand sense You may also need to change the delimiter between genes (i.e. “–” for fusions formatted as such “BCAS4–BCAS3”). The default delimiter is “:”.

pears runs STARsolo to align single-cell reads to a reference genome generating alignment and index files. If you have already have both, you can skip the alignment step in the config file. You should place your alignment files in a directory with the path your_out_dir/STAR/.

Similarly, if you would like to use individual sections of pears you may select/deselect tools. See more details. hyperlink to another section

Adjustable Parameters

The adjustable parameters for pears includes:

• fuscia_mapqual
• fuscia_up/fuscia_down
• flexi_searchlen

parameters	requirements
fuscia_mapqual	the map_qual parameter used by fuscia, this will determine the minimum acceptable MAPQ quality of reads fuscia will look for
fuscia_up/fuscia_down	if there is no corresponding gene annotation in the reference gene.gtf file, you can specify a range (in bp) upstream/downstream of the fusion breakpoint
flexiplex_searchlen	2x the length of sequence (bp) flexiplex will use to search for the fusion sequence. Please note, the longer the search length the longer flexiplex will take to run. See Davidson et al.

Arriba

Link to Arriba

Flexiplex

Link to Flexiplex

Fuscia

Link to Fuscia

Output

Modules and Requirements

Pears includes three modules: alignment, fuscia and flexiplex. You may choose to use any combination of the three, however some may have dependencies.

module	requirements
STAR	reads (randomised*), reference
Arriba	reads, .bam and .bam.bai files
fuscia	reads, list of known fusions (fusion names, chrX:start-end for both genes), .bam and .bam.bai files
flexiplex	reads, list of known fusions (fusion names, cell barcode and UMI length, sequence around the fusion breakpoint)

If your reads are already sorted by coordinate, please randomise before alignmented see issue.

Pears also contains additional scripts to help format and streamline between processes.

gen_masterdata creates the main file fuscia and flexiplex will drawn from. It will output a dataframe with the columns:

• fusion name
• chromosome
• start of the gene (gene1) to site of breakpoint (base1) or site of breakpoint (base2) to end of gene (gene2)
• strand sense
• two sequence +/- n bases around the breakpoint The header of the file includes: fusion_name | chrom1 | gene1 | base1 | sequence1 | strand1 | chrom2 | gene2 | base2 | sequence2 | strand2

Fuscia will create files that include: cell_barcode | molecular_barcode | chrom | start | end .

Flexiplex will create files that include: Read | Cell_barcode | FlankEditDist | BarcodeEditDist | UMI .

format_fuscia and format_flexiplex function to collate the .txt files flexiplex and fuscia output.

gen_report creates the main report file **that is already classified **