Genotype Likelihoods IMputation and PhaSing mEthod

GLIMPSE1 is a phasing and imputation method for large-scale low-coverage sequencing studies.

Main features of the method:

1. Accurate imputed genotype calls. Our method takes advantage of reference panels to produce high quality genotype calls.
2. Accurate phasing. GLIMPSE1 outputs accurate phased haplotypes for the low-coverage sequenced dataset.
3. Low-coverage sequencing outperforms SNP arrays. Imputation using low-coverage sequencing data is competitive to SNP array imputation. Results for European and African-American populations are interactively available on the website.
4. A cost-effective paradigm. GLIMPSE1 realises whole genome imputation from the HRC reference panel for less than 1$.
5. An easy-to-use framework for diploid, haploid and mixed ploidy imputation. Our method is able to perform imputation of samples with different ploidy without the need to split the data.

GLIMPSE1 is available under the MIT licence on the Github repository https://github.com/odelaneau/GLIMPSE.

If you use the GLIMPSE1 in your research work, please cite the following paper:

Simone Rubinacci, Diogo Ribeiro, Robin Hofmeister, Olivier Delaneau. Efficient phasing and imputation of low-coverage sequencing data using large reference panels. Nature Genetics 53.1 (2021): 120-126.

GLIMPSE1 is developed by Simone Rubinacci and Olivier Delaneau at the University of Lausanne, Switzerland.