Harvesting more reads from single-cell combinatorial barcoding data with scarecrow.

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology and Genetics

Background:

• Combinatorial barcoding is crucial for single-cell nucleotide sequencing, enabling scalable tagging of individual cells.
• Technical artifacts can cause positional variability in barcodes, leading to data loss in existing processing tools.
• Current tools often discard reads with truncated or mispositioned barcodes, limiting data recovery.

Purpose of the Study:

• To introduce scarecrow, a novel tool designed to maximize sequencing read retention from single-cell combinatorial barcoding experiments.
• To address the challenge of positional barcode errors ('jitter') in high-throughput sequencing data.
• To improve the accuracy of downstream single-cell analyses by enabling greater data recovery.

Main Methods:

• Scarecrow screens a subset of reads to build position-specific barcode profiles.
• It identifies barcode sequences in each read, accommodating positional errors or 'jitter'.
• Barcode matches are prioritized based on minimizing nucleotide mismatches and jitter, followed by error correction.

Main Results:

• Scarecrow enables flexible identification of barcode sequences despite positional variability.
• The tool effectively accounts for and corrects barcode 'jitter', minimizing data loss.
• Improved data recovery leads to enhanced downstream single-cell analyses.

Conclusions:

• Scarecrow overcomes limitations of existing tools by incorporating 'jitter' into barcode error correction.
• This approach maximizes the retention of sequencing reads from single-cell combinatorial barcoding.
• The open-access Python tool enhances data recovery and downstream single-cell analysis accuracy.