demultiplexing-workflow

This Snakefile pipeline performs demultiplexing starting from Bamfile and VCF file. It will output demultiplexing results for:

• Demuxlet v1
• Demuxlet v2
• SoupOrCell
• Vireo
• SCanSNP

Content overview

• Snakefile Main workflow for demultiplexing
• scripts/Benchmarks_index3.0.0Mod_Plus_Synth.Rmd contains the benchmark script
• Sample* contains demultiplexing results for Demuxlet, Demuxletv2, Soup or Cell, Vireo and SCanSNP
• Sample*/Demultiplexing.report.html contains the result of scripts/Benchmarks_index3.0.0Mod_Plus_Synth.Rmd (sadly no preview is available because of size)

SCanSNP output folder

• Diagnostic plot on DBLs call and LowQuality
• doubletsMarked.tsv with final ID assignment in the format

Barcode	ID_1	ID_2	ID_N	DBL_FirstID_Score	DBL_SecondID_Score	FirstID	SecondID	DropletType	ID	Qual
AAATGC-1	262	2	10	1077.41	39.29	ID1	IDN	Singlet	ID1	GoodQuality
AAAGGC-1	0	450	302	550.15	475.20	ID2	IDN	Doublet	Doublet	Doublet
ACGTCC-1	0	150	1	70.00	50.00	ID2	IDN	Singlet	ID2	LowQuality

In a nutshell:

ID_*: (corrected) reads overlapping ID-private SNPs

ID: ID and doublet information

Qual: Present only if --flagLowQual set to True and contains Quality information (use carefully and looking at mixture fitting QC plot)

Vireo output folder

• Column "donor_id" in file vireoOut/donor_ids.tsv contains informations about type, and assignment

SoC output folder

• Column "status" in file clusters.tsv contains informations about type
• Column "assignment" in file clusters.tsv contains informations about assignment

Demuxlet V1

• Columns "BEST" in file *.best ations about type and assignment e.g: DBL-ID1
• Columns "SNG.1ST"in file *.best ontains informations about assignment assuming singlet type

Demuxlet V2

• Column "DROPLET.TYPE" contains information about type
• Column "BEST.GUESS" contains information about most probable assignment (ordered pair) e.g. IDX,IDY,0.5 (use the first reported id)

R packages