MLL::AF9 degradation induces rapid changes in transcriptional elongation and subsequent loss of an active chromatin landscape.

MLL rearrangements produce fusion oncoproteins that drive leukemia development, but the direct effects of MLL-fusion inactivation remain poorly defined. We designed models with degradable MLL::AF9 where treatment with small molecules induces rapid degradation. We leveraged the kinetics of this system to identify a core subset of MLL::AF9 target genes where MLL::AF9 degradation induces changes in transcriptional elongation within 15 minutes. MLL::AF9 degradation subsequently causes loss of a transcriptionally active chromatin landscape. We used this insight to assess the effectiveness of small molecules that target members of the MLL::AF9 multiprotein complex, specifically DOT1L and MENIN. Combined DOT1L/MENIN inhibition resembles MLL::AF9 degradation, whereas single-agent treatment has more modest effects on MLL::AF9 occupancy and gene expression. Our data show that MLL::AF9 degradation leads to decreases in transcriptional elongation prior to changes in chromatin landscape at select loci and that combined inhibition of chromatin complexes releases the MLL::AF9 oncoprotein from chromatin globally.

Comprehensive single-cell transcriptome lineages of a proto-vertebrate.

Ascidian embryos highlight the importance of cell lineages in animal development. As simple proto-vertebrates, they also provide insights into the evolutionary origins of cell types such as cranial placodes and neural crest cells. Here we have determined single-cell transcriptomes for more than 90,000 cells that span the entirety of development-from the onset of gastrulation to swimming tadpoles-in Ciona intestinalis. Owing to the small numbers of cells in ascidian embryos, this represents an average of over 12-fold coverage for every cell at every stage of development. We used single-cell transcriptome trajectories to construct virtual cell-lineage maps and provisional gene networks for 41 neural subtypes that comprise the larval nervous system. We summarize several applications of these datasets, including annotating the synaptome of swimming tadpoles and tracing the evolutionary origin of cell types such as the vertebrate telencephalon.

Privacy-preserving model evaluation for logistic and linear regression using homomorphically encrypted genotype data.

OBJECTIVE: Linear and logistic regression are widely used statistical techniques in population genetics for analyzing genetic data and uncovering patterns and associations in large genetic datasets, such as identifying genetic variations linked to specific diseases or traits. However, obtaining statistically significant results from these studies requires large amounts of sensitive genotype and phenotype information from thousands of patients, which raises privacy concerns. Although cryptographic techniques such as homomorphic encryption offers a potential solution to the privacy concerns as it allows computations on encrypted data, previous methods leveraging homomorphic encryption have not addressed the confidentiality of shared models, which can leak information about the training data. METHODS: In this work, we present a secure model evaluation method for linear and logistic regression using homomorphic encryption for six prediction tasks, where input genotypes, output phenotypes, and model parameters are all encrypted. RESULTS: Our method ensures no private information leakage during inference and achieves high accuracy (≥93% for all outcomes) with each inference taking less than ten seconds for ∼200 genomes. CONCLUSION: Our study demonstrates that it is possible to perform linear and logistic regression model evaluation while protecting patient confidentiality with theoretical security guarantees. Our implementation and test data are available at https://github.com/G2Lab/privateML/.

VBASS enables integration of single cell gene expression data in Bayesian association analysis of rare variants.

Rare or de novo variants have substantial contribution to human diseases, but the statistical power to identify risk genes by rare variants is generally low due to rarity of genotype data. Previous studies have shown that risk genes usually have high expression in relevant cell types, although for many conditions the identity of these cell types are largely unknown. Recent efforts in single cell atlas in human and model organisms produced large amount of gene expression data. Here we present VBASS, a Bayesian method that integrates single-cell expression and de novo variant (DNV) data to improve power of disease risk gene discovery. VBASS models disease risk prior as a function of expression profiles, approximated by deep neural networks. It learns the weights of neural networks and parameters of Gamma-Poisson likelihood models of DNV counts jointly from expression and genetics data. On simulated data, VBASS shows proper error rate control and better power than state-of-the-art methods. We applied VBASS to published datasets and identified more candidate risk genes with supports from literature or data from independent cohorts. VBASS can be generalized to integrate other types of functional genomics data in statistical genetics analysis.

Hierarchical Concept Relations Improve Detection of Off-Label Drug Use in Electronic Health Records Data.

Real-world clinical practice commonly veers from formal drug approvals in off-label use, accounting for 21% of prescriptions for common drugs. Due to its ad hoc nature, off-label use typically goes undocumented, evading the safety and efficacy scrutiny of clinical trials. A systematic and automated approach to detection of these uses in the electronic health record (EHR) would enable improved safety monitoring, provide insight into prescribing patterns, and support real-world evidence appraisal. Domain knowledge provided by medication-indication knowledge bases has been shown to improve the accuracy of EHR-based automated detection of off-label use, but remains limited due to diverse concept representations and granularities across data sources. We present a method to leverage hierarchical concept knowledge to align medication-indication knowledge with EHR data for automated detection of off-label drug use in clinical practice. We demonstrate an over two-fold increase in detected off-label diagnoses when leveraging hierarchical knowledge relative to direct concept matching alone.