Thank you for your interest in MACIE. MACIE (Multi-dimensional Annotation Class Integrative Estimation) is an unsupervised multivariate mixed model framework to assess multi-dimensional functional impacts for both coding and non-coding variants in the human genome. MACIE integrates a variety of functional annotations, including protein function scores, evolutionary conservation scores, and epigenetic annotations from ENCODE and Roadmap Epigenomics, and estimates the joint posterior probabilities of each genetic variant being functional.

The MACIE scores (and other integrative scores) used in all benchmarking examples are available for download here. Precomputed MACIE scores for every possible variant in the human genome are available for download.

The code used for training MACIE models are available here.

All genomic coordinates are given in NCBI Build 37/UCSC hg19.

Xihao Li*, Godwin Yung*, Hufeng Zhou, Ryan Sun, Zilin Li, Yaowu Liu, Iuliana Ionita-Laza, and Xihong Lin (2021+) "A Multi-Dimensional Integrative Scoring Framework for Predicting Functional Variants in the Human Genome". Submitted.