A New Genetic Tool for Predicting Regulatory Sequences

 

    A project conducted by members of the UC San Diego Department of Medicine resulted in the creation of EUGENe, a tool designed to assist in the prediction of regulatory sequences. The team lists a lack of deep-learning methods for regulatory genetics that follow the FAIR(findable, accessible, interoperable, reusable) principle, and created EUGENe to fill that gap in technology.

    The team states that EUGENe is  "a simple, flexible and extensible interface for streamlining and customizing end-to-end deep-learning sequence analyses." The team discusses the lack of software that can effectively tackle the problems faced in deep-learning related genetics and goes into detail about the failures that each of the current major packages has in regards to a comprehensive package that can handle an end-to-end workflow while integrating with other available packages.

Article on EUGENe

Klie, A., Laub, D., Talwar, J.V. et al. Predictive analyses of regulatory sequences with EUGENe. Nat Comput Sci 3, 946–956 (2023). https://doi.org/10.1038/s43588-023-00544-w

Article on Janggu, a major software package that handles deep-learning in genomics

Kopp, W., Monti, R., Tamburrini, A. et al. Deep learning for genomics using Janggu. Nat Commun 11, 3488 (2020). https://doi.org/10.1038/s41467-020-17155-y

Using Artificial Intelligence to Impute Phenotypes on Population Scale Databases

 

    A report from UCLA details the usage of analyzing large scale databases to impute phenotypes in databases that lack data on phenotypes. Phenotypes in these databases are often missing in many of the individuals in the databases, limiting the utility that these databases have. The group created a new imputation method for phenotypes based off of the UK Biobank dataset. The accuracy of their imputation method, dubbed AutoComplete, outperformed the next best method, called SoftImpute, with the greatest improvements being in the psychiatric and cardiovascular phenotypes. 

    The significance of their imputer is centered around the fact that they take into consideration not only the relationships between phenotypes and the data available, but also the metadata of the database used. They took into account the "patterns of missingness" in the data, which are often structured due to the way that the data is gathered and placed into the databases.

    This report not only exemplifies the importance of having strong computational methods in genetics but also the significance of properly reporting data. Datasets that contain complete information not only make it easier to work with the data, but also provide more accurate results once the data is processed.

The Article

An, U., Pazokitoroudi, A., Alvarez, M. et al. Deep learning-based phenotype imputation on population-scale biobank data increases genetic discoveries. Nat Genet (2023). https://doi.org/10.1038/s41588-023-01558-w

Reference to SoftImpute

Hastie, T., Mazumder, R., Lee, J. D., & Zadeh, R. (2015). Matrix completion and low-rank SVD via fast alternating least squares. The Journal of Machine Learning Research, 16(1), 3367-3402.