Alan M Moses

The paper shows that Pericentrin contains an internal sequence that directly activates dynein for co-translational transport, challenging the notion that dynein relies solely on N-terminal adaptors.

The paper uses in vivo parallelized reporter assays in C. elegans to identify tissue-specific splicing regulatory sequences and their functional relevance.

The paper maps the functional landscape of the human intrinsically disordered proteome using evolutionary signatures to classify and predict IDR functions.

This paper introduces EnhancAR, a deep learning model that generates functionally diverse enhancers using evolutionary homologs instead of cell type labels.

This study identifies PRRC2A, PRRC2B, and PRRC2C as stress granule proteins that regulate translation by interacting with the eIF3 complex and 48S PIC factors.

This paper introduces a pretraining task for genome language models that predicts evolutionary rate, showing improved performance over sequence-only pretraining.

AlphaFlex generates atomistic conformational ensembles for human proteins with intrinsically disordered regions, integrating AlphaFold models and implicit bilayers for transmembrane proteins.

The paper introduces a zero-shot protein segmentation method using ProtT5 embeddings to identify and categorize functional protein segments without sequence conservation reliance.

The paper infers fungal cis-regulatory networks from genome sequences using unsupervised and interpretable representation learning, identifying conserved motif signals that predict gene expression and regulation.

This paper introduces a framework to interpret attention mechanisms in genomic transformer models by correlating attention scores with biological annotations, showing how biological features are learned through pre-training and fine-tuning.

This paper evaluates how background cells affect single-cell microscopy models and finds they reduce localization classification accuracy by up to 15.8%.

This paper introduces a self-supervised whole-genome representation learning method for bacteria, showing it outperforms traditional feature-based approaches in predicting phenotypes.

This study identifies functional relationships between TSC22D, WNK, and NRBP gene families in macromolecular crowding sensing and cell volume regulation across metazoans.

This paper improves supervised deep learning for regulatory genomics by using phylogenetic augmentation to enhance model performance on small datasets.

The paper designs synthetic intrinsically disordered regions by minimizing molecular property distances to known IDRs, showing expected patterns in cells.

The paper defines a feature space for non-coding DNA sequences using phylogenetic average motif scores, revealing functional similarity among promoters and 3' UTRs.

This study maps RNA translation dynamics in the human brain using ribosome profiling, identifying thousands of novel translation events and brain-specific microproteins linked to schizophrenia.

The paper introduces a method to discover molecular features of intrinsically disordered regions using contrastive learning with evolutionary conservation.

The paper identifies a recurrent SHANK3 frameshift variant in Autism Spectrum Disorder, showing de novo mutations and variable expressivity.

Architect improves metabolic model quality by combining enzyme annotations from multiple tools and gap-filling based on likelihood scores.

The paper identifies a diversity threshold of transcription factor binding sites (TFBS) required for enhancer activity in embryonic stem cells, showing that a diverse set of TFBS can replace optimized regulatory sequences.

CPCProt pretrains protein embeddings by telling a protein's own next fragment from fragments of random proteins, reportedly matching benchmarked self-supervised models with 2-10% of their parameters.

This paper introduces a self-supervised CNN method to learn high-quality features for single-cell microscopy images without labeled data, showing improved performance in protein localization and analysis.

The paper identifies evolutionary signatures in intrinsically disordered regions that correlate with functional annotations and phenotypes.

The paper introduces a semi-supervised clustering model for identifying active enhancers in mouse heart tissues using heterogeneous genomic data.

The paper identifies functional short linear motifs in disordered regions of HOG signaling proteins and shows their quantitative impact on pathway output through experimental and computational approaches.

The paper shows that constant calcium stress causes stochastic nuclear localization pulses of Crz1 in yeast, proposing a model where time delays and stochasticity convert analog calcium signals into digital pulses.