Biology subjects

Coffey, R.

Publications and source records attributed to Coffey, R..

7 recordsLinked to original sources

bioRxiv2025

Systematic feature and architecture evaluation reveals tokenized learned embeddings enhance siRNA efficacy prediction

Recent advances in machine learning have improved the prediction of siRNA efficacy, with graph neural networks and transformer-based encodings leading the way. However, existing models still face challenges, including potential inaccuracies in thermodynamic feature calculations (such as incorrect strand selection for siRNA-mRNA Gibbs free energy), limited effective utilization of available datasets, and a lack of systematic model refinement. In this study, I systematically evaluated the predictive power of individual features and neural network architectures to identify the most effective configurations. This process led to the development of RN.Ai-Predict, a model built upon a tokenized learned embedding for nucleotide sequences. This work demonstrates that a methodical approach to feature selection and hyperparameter tuning, particularly favoring learned embeddings, can yield a more accurate and reliable model for predicting siRNA efficacy, outperforming more complex architectures in generalizability.

bioinformatics↗

bioRxiv2025

Reinforced molecular dynamics: Physics-infused generative machine learning model explores CRBN activation process

We propose a simple and practical machine learning-based desktop solution for modeling biologically relevant protein motions. We termed our technology reinforced molecular dynamics (rMD) combining MD trajectory data and free-energy (FE) map data to train a dual-loss function autoencoder network that can explore conformational space more efficiently than the underlying MD simulation. The key insight of rMD is that it effectively replaces the latent space with an FE map, thus infusing the autoencoder network with a physical context. The FE map is computed from an MD simulation over a low-dimensional collective variable space that captures some biological function. One can directly use then the FE map for example, to generate more protein structures in poorly sampled regions, follow paths on the FE map to explore conformational transitions, etc. The rMD technology is entirely self-contained, does not rely on any pre-trained model, and can be run on a single GPU desktop computer. We present our rMD computations in a key area of molecular-glue degraders aimed at a deeper understanding of the structural transition from open to closed conformations of CRBN.

molecular biology↗

bioRxiv2022

Bulk and Mosaic Deletions of Egfr Reveal Regionally Defined Gliogenesis in the Developing Mouse Forebrain

The Epidermal growth factor receptor (EGFR) plays a role in cell proliferation and differentiation during healthy development and tumor growth, however its requirement for brain development remains unclear. Here we used a conditional mouse allele for Egfr to examine its contributions to perinatal forebrain development at the tissue level. Subtractive bulk ventral and dorsal forebrain deletions of Egfr uncovered significant and permanent decreases in oligodendrogenesis and myelination in the cortex and corpus callosum. Additionally, an increase in astrogenesis or reactive astrocytes in effected regions was evident in response to cortical scarring. Sparse deletion using Mosaic Analysis with Double Markers (MADM) surprisingly revealed a regional requirement for EGFR in rostrodorsal, but not ventrocaudal glial lineages including both astrocytes and oligodendrocytes. The EGFR-independent ventral glial progenitors may compensate for the missing EGFR-dependent dorsal glia in the bulk Egfr-deleted forebrain, potentially exposing a regenerative population of gliogenic progenitors in the mouse forebrain.

neuroscience↗

bioRxiv2022

Modeling the effect of cooperativity in ternary complex formation and targeted protein degradation mediated by heterobifunctional degraders

Chemically induced proximity between certain endogenous enzymes and a protein of interest (POI) inside cells may cause post-translational modifications to the POI with biological consequences and potential therapeutic effects. Heterobifunctional (HBF) molecules that bind with one functional part to a target POI and with the other to an E3 ligase induce the formation of a target-HBF-E3 ternary complex, which can lead to ubiquitination and proteasomal degradation of the POI. Targeted protein degra-dation (TPD) by HBFs offers a promising approach to modulating disease-associated proteins, especially those that are intractable using other therapeutic approaches, such as enzymatic inhibition. The three-way interactions among the HBF, the target POI, and the ligase--including the protein-protein interaction (PPI) between the POI and the ligase--contribute to the stability of the ternary complex, manifested as positive or negative binding cooperativity in its formation. How such cooperativity affects HBF-mediated degradation is an open question. In this work, we develop a pharmaco-dynamic model that describes the kinetics of the key reactions in the TPD process, and we use this model to investigate the role of cooperativity in the ternary complex formation and in the target POI degradation. Our model predicts that, under certain conditions, increasing cooperativity may diminish degradation, implying an optimal range of cooperativity values for efficient degradation. We also develop a statistical inference model for determining cooperativity in intracellular ternary complex formation from cellular assay data, and demonstrate it by quantifying the change in cooperativity due to site-directed mutagenesis at the POI-ligase interface of the SMARCA2-ACBI1-VHL ternary complex. Our pharmacodynamic model provides a quantitative framework to dissect the complex HBF-mediated TPD process and may inform the rational design of effective HBF degraders.

biochemistry↗

bioRxiv2021

Formation and recycling of an active epigenetic mark mediated by cell cycle-specific RNAs

The mechanisms by which epigenetic modifications are established in gene regulatory regions of active genes remain poorly understood. The data presented show that the establishment and recycling of a major epigenetic mark, the acetylated form of the replacement histone H2A.Z, is regulated by cell cycle-specific long noncoding RNAs encoded in regions adjacent to the promoters of active genes. These transcripts, termed SPEARs (S Phase EArly RNAs), are induced in early S phase: their expression precedes that of the downstream genes on which they exert their regulatory action. SPEARs drive the modification and deposition of the acetylated form of histone H2A.Z by bringing together the replacement histone and the histone acetyl transferase TIP60. This widespread bimodal pathway constitutes a novel RNA-mediated mechanism for the establishment of epigenetic marks and cell-specific epigenetic profiles, thereby providing a unifying explanation for the accuracy and persistence of epigenetic marks on chromatin.

genetics↗

bioRxiv2021

Erythroid overproduction of erythroferrone causes iron overload and developmental abnormalities in mice

The hormone erythroferrone (ERFE) is produced by erythroid cells in response to hemorrhage, hypoxia or other erythropoietic stimuli, and suppresses the hepatic production of the iron-regulatory hormone hepcidin, thereby mobilizing iron for erythropoiesis. Suppression of hepcidin by ERFE is thought to be mediated by interference with paracrine BMP signaling that regulates hepcidin transcription in hepatocytes. In anemias with ineffective erythropoiesis, ERFE is pathologically overproduced but its contribution to the clinical manifestations of these anemias is not well understood. We generated three lines of transgenic mice with graded erythroid overexpression of ERFE and showed that they developed dose-dependent iron overload, impaired hepatic BMP signaling and relative hepcidin deficiency. These findings add to the evidence that ERFE is a mediator of iron overload in conditions where ERFE is overproduced, including anemias with ineffective erythropoiesis. At the highest levels of ERFE overexpression the mice manifested decreased perinatal survival, impaired growth, small hypofunctional kidneys, decreased gonadal fat depots and neurobehavioral abnormalities, all consistent with impaired organ-specific BMP signaling during development. Neutralizing excessive ERFE in congenital anemias with ineffective erythropoiesis may not only prevent iron overload but may have additional benefits for growth and development. Key PointsO_LIChronic erythroid overproduction of erythroferrone dose-dependently suppresses hepcidin, causing iron overload even in the absence of anemia C_LIO_LIHigh level overexpression of erythroferrone can cause delayed growth, impaired kidney function and other developmental abnormalities consistent with altered BMP signaling C_LI

physiology↗

bioRxiv2020

Myeloid lncRNA LOUP Mediates Opposing Regulatory Effects of RUNX1 and RUNX1-ETO in t(8;21) AML

The mechanism underlying cell type-specific gene induction conferred by ubiquitous transcription factors as well as disruptions caused by their chimeric derivatives in leukemia is not well understood. Here we investigate whether RNAs coordinate with transcription factors to drive myeloid gene transcription. In an integrated genome-wide approach surveying for gene loci exhibiting concurrent RNA- and DNA-interactions with the broadly expressed transcription factor RUNX1, we identified the long noncoding RNA LOUP. This myeloid-specific and polyadenylated lncRNA induces myeloid differentiation and inhibits cell growth, acting as a transcriptional inducer of the myeloid master regulator PU.1. Mechanistically, LOUP recruits RUNX1 to both the PU.1 enhancer and the promoter, leading to the formation of an active chromatin loop. In t(8;21) acute myeloid leukemia, wherein RUNX1 is fused to ETO, the resulting oncogenic fusion protein RUNX1-ETO limits chromatin accessibility at the LOUP locus, causing inhibition of LOUP and PU.1 expression. These findings highlight the important role of the interplay between cell type-specific RNAs and transcription factors as well as their oncogenic derivatives in modulating lineage-gene activation and raise the possibility that RNA regulators of transcription factors represent alternative targets for therapeutic development. KEY POINTSO_LIlncRNA LOUP coordinates with RUNX1 to induces PU.1 long-range transcription, conferring myeloid differentiation and inhibiting cell growth. C_LIO_LIRUNX1-ETO limits chromatin accessibility at the LOUP locus, causing inhibition of LOUP and PU.1 expression in t(8;21) AML. C_LI

cancer biology↗