BaySeq-Y

Direct probabilistic quantification of mosaic loss of chromosome Y from sequencing data

BaySeq-Y is a computational framework for estimating mosaic loss of chromosome Y (LOY) from DNA sequencing data, including whole-exome sequencing (WES) and whole-genome sequencing (WGS). BaySeq-Y directly estimates the fraction of cells lacking chromosome Y using a Bayesian framework that integrates multiple LOY-associated genomic signals.

This repository contains three R implementations of BaySeq-Y for different input data types:

• BaySeq-Y_no_phasing.r
  LOY estimation using read depth (DP) in the Y non-pseudoautosomal region (Y-non-PAR) and allelic depth (AD) in the pseudoautosomal region (PAR).

• BaySeq-Y_phasing.r
  LOY estimation using phased allelic depth (AD) information in PAR1.

• BaySeq-Y_phasing_fast.r
  Fast implementation of LOY estimation using phased allelic depth information in PAR1.

Requirements

• R (version 4.0 or later recommended)
• Required R packages:
  • rjags
  • coda
  • data.table
  • dplyr

Additional package requirements may depend on the implementation used.

Input Files

BaySeq-Y_no_phasing.r

Required inputs:

1. DP in Y-non-PAR
2. AD in PAR
3. Median DP in X-non-PAR
4. Median BAF in autosomes

Example input files:

• DP_sample_1.tsv
• AD_sample_1.tsv

BaySeq-Y_phasing.r and BaySeq-Y_phasing_fast.r

Required inputs:

1. Phased AD in PAR1
2. Number of blocks (recommended: 10–20)
3. Median BAF in autosomes

Example input file:

• Phased_AD_sample_1.tsv

Usage

LOY estimation using Y-non-PAR and PAR data

Rscript BaySeq-Y_no_phasing.r \
    DP_sample_1.tsv \
    AD_sample_1.tsv \
    19 \
    0.5 \
    <result_file_name> \
    <sample_ID>

Arguments:

1. DP input file
2. AD input file
3. Median DP in X-non-PAR
4. Median BAF in autosomes
5. Output file name
6. Sample ID

LOY estimation using phased PAR1 data

Rscript BaySeq-Y_phasing.r \
    Phased_AD_sample_1.tsv \
    20 \
    0.5 \
    <result_file_name> \
    <sample_ID>

or

Rscript BaySeq-Y_phasing_fast.r \
    Phased_AD_sample_1.tsv \
    20 \
    0.5 \
    <result_file_name> \
    <sample_ID>

Arguments:

1. Phased AD input file
2. Number of blocks
3. Median BAF in autosomes
4. Output file name
5. Sample ID

Output

Each script generates an output file containing:

• Sample ID
• Estimated LOY fraction
• Lower bound of the 95% confidence interval
• Upper bound of the 95% confidence interval

Additionally:

• BaySeq-Y_no_phasing.r reports the ΔBIC statistic used to assess evidence for LOY.

Interpretation

The reported LOY estimate represents the estimated fraction of cells lacking chromosome Y in the analyzed sample.

Typical interpretation:

LOY Estimate	Interpretation
0.00	No detectable LOY
0.10	Approximately 10% of cells have lost chromosome Y
0.25	Approximately 25% of cells have lost chromosome Y
0.50	Approximately 50% of cells have lost chromosome Y

Confidence intervals provide uncertainty estimates around the inferred LOY fraction.

Citation

If you use BaySeq-Y in your research, please cite:

Lin J-R, Chang Y-C, Maslov AY, Song Y, Gao T, Shan J, Bennett D, Milman S, Barzilai N, Vijg J, Montagna C, Zhang ZD
Direct probabilistic quantification of mosaic loss of chromosome Y from sequencing data.
bioRxiv 2026.06.26.734767; doi: https://doi.org/10.64898/2026.06.26.734767

Contact

Zhengdong D. Zhang
Department of Genetics
Albert Einstein College of Medicine

For questions regarding BaySeq-Y, please contact:

zhengdong.zhang@einsteinmed.edu