Biology subjects

Moses, A. M.

Publications and source records attributed to Moses, A. M..

At least 19 recordsLinked to original sources

bioRxiv2026

Activating adaptor-like sequences in pericentrin mediate its transport by dynein

Right before a cell divides, the centrosome rapidly increases its size and microtubule organizing activity through a process termed centrosome maturation. PCNT, a coiled-coil centrosomal protein, is synthesized and transported to the centrosome simultaneously by dynein. This dynein-mediated co-translational transport of PCNT facilitates centrosome maturation and mitotic spindle formation. How dynein engages and transports PCNT, however, remains unclear. Here we find that PCNT residues 1393-1525 are required for mediating dynein-dynactin interaction. PCNT (1393-1525) also shares similar sequence features with canonical dynein cargo adaptors. PCNT is predicted to interact with dynein heavy chains through similar contact sites used by canonical dynein cargo adaptors. Introducing point mutations at those contact sites abolishes dyneinmediated transport in a peroxisome motility assay. Our results suggest that PCNT contains adaptorlike sequences to bind and activate dynein directly during its transport to the centrosome. Our data also suggest that dynein may engage with some of its cargoes directly without an adaptor.

cell biology↗

bioRxiv2026

Evolutionary conditioning enables guided generation of functionally diverse enhancers

Deep learning has been instrumental in our understanding of how enhancers encode regulatory information in their DNA sequence and has demonstrated preliminary success with enhancer design. However, the prevailing approach for enhancer design, cell type label conditioning, depends on labeled data from massively parallel reporter assays, which only exists for a handful of cell types. We propose EnhancAR, an autoregressive model trained on sets of unaligned homologous enhancer sequences to learn the function of the enhancer conserved over evolution and generate sequences that resemble real homologs. By training EnhancAR on 1.7 million human enhancer homolog sets spanning 1,888 cell types, EnhancAR generates enhancers for a variety of contexts without being conditioned on a cell type label. We computationally validate that when conditioned on a set of enhancer homologs, EnhancAR generates novel and diverse sequences that preserve the functional properties of the homologs. By prompting EnhancAR with homologs for existing cell type specific enhancers, we design enhancers with similar predicted cell type specificity. We further demonstrate that when trained on length sorted homologs, EnhancAR can design enhancers shorter than the conditioning homologs that preserve the predicted activity. In summary, we find that leveraging evolutionary information in enhancer homologs enables a more flexible and general paradigm for designing enhancers with specific functions.

bioengineering↗

bioRxiv2025

AlphaFlex: Ensembles of the human proteome representing disordered regions

More than two thirds of proteins in the human proteome are predicted to contain intrinsically disordered regions (IDRs), which lack stable folded structure. IDRs are critical for biological regulation and organization, as targets for post-translational modifications, and as mediators of biomolecular condensates. To address the pressing need for better structural models enabling functional insight, we developed AlphaFlex to model fully atomistic conformer ensembles for proteins predicted to have IDRs, modeled in the context of AlphaFold folded domains and an implicit bilayer for transmembrane proteins. The AlphaFlex resource provides conformational ensembles of human proteins from the AlphaFold database with identified IDRs in the Protein Ensemble Database that is mirrored in UniProt. This transformative resource of AlphaFlex ensembles provides physically and biologically relevant full-length models for IDR proteins, including scaffold proteins, those with IDR:folded-domain interactions, regulatory and condensate proteins requiring exposed binding elements, conditionally folding IDRs, and transmembrane proteins containing IDRs.

biochemistry↗

bioRxiv2025

An in vivo parallelized reporter assay to uncover tissue-specific splicing regulatory sequences in a multicellular animal

Introns play a critical role in regulating alternative splicing. However, identifying functional intronic motifs is challenging due to their short and degenerate sequence composition. Massively parallel reporter assays have provided insights into cis-regulatory logic governing alternative splicing, but these approaches are generally performed in cell culture, limiting their ability to capture tissue-specific contexts. Here, we implemented in vivo Parallelized Reporter Assays in C. elegans neurons and muscle cells to screen for intronic enhancer and silencer motifs among thousands of randomized sequences. We identified nearly 200 sequences regulating splicing in these tissues. We uncovered core sub-sequences with tissue-biased enhancing and silencing activity, including motifs recognized by well-characterized RNA-binding proteins, and orphan motifs with no obvious cognate binding protein. Mapping our PRA-derived motifs to native introns flanking tissue-biased alternative exons revealed their conservation across nematodes, supporting their functional relevance. Additionally, individual intronic regions frequently contained diverse combinations of these motifs, indicative of complex engagement of these sequences by RNA-binding proteins. Finally, we performed targeted mutagenesis of PRA-derived intronic enhancers flanking a neuronal microexon, identifying key cis-regulatory determinants of microexon splicing. Together, our study provides a framework to explore the role of intronic regions in tissue-specific splicing regulation within a multicellular organism. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=107 SRC="FIGDIR/small/670696v1_ufig1.gif" ALT="Figure 1"> View larger version (24K): org.highwire.dtl.DTLVardef@84d5daorg.highwire.dtl.DTLVardef@f3f625org.highwire.dtl.DTLVardef@a53739org.highwire.dtl.DTLVardef@ccbb03_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2025

Teneurin-3 and latrophilin-2 are required for somatotopic map formation and somatosensory topognosis

Somatotopy is a recurring organisational feature of the somatosensory system where adjacent neurons and their connections represent adjacent regions of the body. The molecular mechanisms governing the formation of such "body maps" remain largely unknown. Here we demonstrate that the cell surface proteins teneurin-3 and latrophilin-2 are expressed in opposing gradients aligning with the somatotopic map in the dorsal horn of the mouse spinal cord. Genetic manipulation of these proteins in spinal dorsal horn or sensory neurons distorts the somatotopy of neuronal connections and impairs accurate localisation of a noxious stimulus on the surface of the body. Our work provides the foundation for a molecular model of somatotopic map formation and insights into the role of somatotopic maps in the ability to accurately locate somatosensory stimuli or topognosis.

neuroscience↗

bioRxiv2025

Unsupervised Whole-Genome Representation Learning Captures Bacterial Phenotypes

Shifting from hand-crafted to learned representations of data has revolutionized fields like natural language processing and computer vision. Despite this, current approaches to bacterial phenotype prediction from the genome rely on training machine learning models on hand-crafted features, often binary indicators or counts of the presence of different conserved genomic elements and protein domains. Defining these shared elements and domains as our "genomic element vocabulary", we tokenize entire bacterial genomes as sequences of these conserved elements and take advantage of advances in long-context language modeling to perform self-supervised whole-genome representation learning (WGRL). Through multi-task pretraining on a phylogenetically diverse dataset of hundreds of thousands of bacterial genomes, we present a genomic language model which produces representations of input genomes with features predictive of a broad range of phenotypes. We assess the quality of the learned representations through k-nearest neighbours prediction of 25 bacterial phenotypes, finding our WGRL representations more predictive than standard protein domain presence/absence representations for 23/25 different phenotypes. We additionally find the WGRL representations are robust to both poor genome assembly quality and incompleteness. Through learning the relationships between evolutionarily conserved genomic elements with self-supervised long-context language modeling, we demonstrate the first approach for extracting general-purpose whole-genome representations while preserving gene order.

bioinformatics↗

bioRxiv2025

Zero-shot segmentation using embeddings from a protein language model identifies functional regions in the human proteome

The biological function of a protein is often determined by its distinct functional units, such as folded domains and intrinsically disordered regions. Identifying and categorizing these protein segments from sequence has been a major focus in computational biology which has enabled the automatic annotation of folded protein domains. Here we show that embeddings from the unsupervised protein language model ProtT5 can be used to identify and categorize protein segments without relying on conserved patterns in primary amino acid sequence. We present Zero-shot Protein Segmentation (ZPS), where we use embeddings from ProtT5 to predict the boundaries of protein segments without training or fine-tuning any parameters. We find that ZPS boundary predictions for the human proteome are more consistent with reviewed annotations from UniProt than established bioinformatics tools and ProtT5 embeddings of ZPS segments can categorize folded domains, sub-domains, and intrinsically disordered regions. To explore ZPS predictions, we introduce a new way to visualize protein embeddings that closely resembles diagrams of distinct functional units in protein biology. Since ZPS and segment embeddings can be used without training or fine-tuning, the approach is not biased towards known annotations and can used to identify and categorize unannotated protein segments. We used the segment embeddings to identify unannotated mitochondrion targeting signals and SYGQ-rich prion-like domains, which are functional regions within intrinsically disordered regions. We expect the protein segment organization revealed here to lead to valuable information about protein function, including about intrinsically disordered regions and other less understood protein regions.

bioinformatics↗

bioRxiv2025

Inferring fungal cis-regulatory networks from genome sequences

Gene expression patterns are determined to a large extent by transcription factor binding to non-coding regulatory regions in the genome. However, gene expression cannot yet be systematically predicted from genome sequences, in part because non-functional matches to the sequence patterns (motifs) recognized by transcription factors (TFs) occur frequently throughout the genome. Large-scale functional genomics data for many TFs has enabled characterization of regulatory networks in experimentally accessible cells such as budding yeast. Beyond yeast, fungi are important industrial organisms and pathogens, but large-scale functional data is only sporadically available. Uncharacterized regulatory networks control key pathways and gene expression programs associated with fungal phenotypes. Here we explore a sequence-only approach to inferring regulatory networks by leveraging the 100s of genomes now available for many clades of fungi. We use gene orthology as the learning signal to infer interpretable, TF motif-based representations of non-coding regulatory regions. Using these representations to identify conserved signals for motifs, comparative genomics can be scaled to evolutionary comparisons where sequence similarity cannot be detected. We show that similarity of these conserved motif signals predicts gene expression and regulation better than using experimental data, and that we can infer known and novel regulatory connections in diverse fungi. Our new predictions include a pathway for recombination in C. albicans and pathways for mating and an RNAi immune response in Neurospora. Taken together, our results indicate that specific hypotheses about transcriptional regulation in fungi can be obtained for many genes from genome sequence analysis alone.

systems biology↗

bioRxiv2024

A Functional Map of the Human Intrinsically Disordered Proteome

Intrinsically disordered regions (IDRs) represent at least one-third of the human proteome and defy the established structure-function paradigm. Because IDRs often have limited positional sequence conservation, the functional classification of IDRs using standard bioinformatics is generally not possible. Here, we show that evolutionarily conserved molecular features of the intrinsically disordered human proteome (IDR-ome), termed evolutionary signatures, enable classification and prediction of IDR functions. Hierarchical clustering of the human IDR-ome based on evolutionary signatures reveals strong enrichments for frequently studied functions of IDRs in transcription and RNA processing, as well as diverse, rarely studied functions, ranging from sub-cellular localization and biomolecular condensates to cellular signaling, transmembrane transport, and the constitution of the cytoskeleton. We exploit the information that is encoded within evolutionary conservation of molecular features to propose functional annotations for every IDR in the human proteome, inspect the conserved molecular features that correlate with different functions, and discover frequently co-occurring IDR functions on the proteome scale. Further, we identify patterns of evolutionary conserved molecular features of IDRs within proteins of unknown function and disease-risk genes for conditions such as cancer and developmental disorders. Our map of the human IDR-ome should be a valuable resource that aids in the discovery of new IDR biology.

bioinformatics↗

bioRxiv2023

Improving the performance of supervised deep learning for regulatory genomics using phylogenetic augmentation

Structured abstractO_ST_ABSMotivationC_ST_ABSSupervised deep learning is used to model the complex relationship between genomic sequence and regulatory function. Understanding how these models make predictions can provide biological insight into regulatory functions. Given the complexity of the sequence to regulatory function mapping (the cis-regulatory code), it has been suggested that the genome contains insufficient sequence variation to train models with suitable complexity. Data augmentation is a widely used approach to increase the data variation available for model training, however current data augmentation methods for genomic sequence data are limited. ResultsInspired by the success of comparative genomics, we show that augmenting genomic sequences with evolutionarily related sequences from other species, which we term phylogenetic augmentation, improves the performance of deep learning models trained on regulatory genomic sequences to predict high-throughput functional assay measurements. Additionally, we show that phylogenetic augmentation can rescue model performance when the training set is down-sampled and permits deep learning on a real-world small dataset, demonstrating that this approach improves experimental data efficiency. Overall, this data augmentation method represents a solution for improving model performance that is applicable to many supervised deep learning problems in genomics. Availability and implementationThe open-source GitHub repository agduncan94/phylogenetic_augmentation_paper includes the code for rerunning the analyses here and recreating the figures. Contactalan.moses@utoronto.ca

bioinformatics↗

bioRxiv2023

Computational design of intrinsically disordered protein regions by matching bulk molecular properties

Algorithms for computational protein design usually begin with a 3D structure and design an amino acid sequence that will fold into that structure. Since intrinsically disordered protein regions lack stable 3D structures, strategies for their design have been limited. Here we describe and validate a general computational strategy for design of intrinsically disordered regions (IDRs). Our algorithm designs synthetic IDRs by minimizing the distance of an initially random amino acid sequence to a known IDR in a high-dimensional space of molecular properties, such as repeat content, charge and hydrophobicity. We tested a handful of IDRs designed to target proteins to mitochondria and heat-induced condensates, and several showed the expected patterns in cells. One sentence summaryWe design synthetic mitochondrial targeting signals and condensate forming intrinsically disordered regions

synthetic biology↗

bioRxiv2023

Functional similarity of non-coding regions is revealed in phylogenetic average motif score representations

Here we frame the cis-regulatory code (that connects the regulatory functions of non-coding regions, such as promoters and UTRs, to their DNA sequences) as a representation building problem. Representation learning has emerged as a new approach to understand function of DNA and proteins, by projecting sequences into high-dimensional feature spaces, where the features are learned from data by a neural network. Inspired by these approaches, we seek to define a feature space where non-coding regions with similar regulatory functions are nearby each other. As a first attempt, we engineered features based on matches to biochemically characterized regulatory motifs in the DNA sequences of non-coding regions. Remarkably, we found that functionally similar promoters and 3 UTRs could be grouped together in a feature space defined by simple averages of the best match scores in (unaligned) orthologous non-coding regions, which we refer to as phylogenetic average motif scores. Perhaps most important, because this feature space is based on known motifs and not fit to any data, it is fully interpretable and not limited to any particular cell type or experimental context. We find that we can read off known regulatory relationships and evolutionary rewiring from visualizations of phylogenetic average motif score representations, and that predicted regulatory interactions based on neighbors in the feature space are borne out in transcription factor deletion experiments. Phylogenetic averages of match scores to known motifs is a baseline for representation learning applied to non-coding sequences, and may continue to improve as databases of motifs become more complete.

genomics↗

bioRxiv2022

Systematic identification of conditionally folded intrinsically disordered regions by AlphaFold2

The AlphaFold Protein Structure Database contains predicted structures for millions of proteins. For the majority of human proteins that contain intrinsically disordered regions (IDRs), which do not adopt a stable structure, it is generally assumed these regions have low AlphaFold2 confidence scores that reflect low-confidence structural predictions. Here, we show that AlphaFold2 assigns confident structures to nearly 15% of human IDRs. By comparison to experimental NMR data for a subset of IDRs that are known to conditionally fold (i.e., upon binding or under other specific conditions), we find that AlphaFold2 often predicts the structure of the conditionally folded state. Based on databases of IDRs that are known to conditionally fold, we estimate that AlphaFold2 can identify conditionally folding IDRs at a precision as high as 88% at a 10% false positive rate, which is remarkable considering that conditionally folded IDR structures were minimally represented in its training data. We find that human disease mutations are nearly 5-fold enriched in conditionally folded IDRs over IDRs in general, and that up to 80% of IDRs in prokaryotes are predicted to conditionally fold, compared to less than 20% of eukaryotic IDRs. These results indicate that a large majority of IDRs in the proteomes of human and other eukaryotes function in the absence of conditional folding, but the regions that do acquire folds are more sensitive to mutations. We emphasize that the AlphaFold2 predictions do not reveal functionally relevant structural plasticity within IDRs and cannot offer realistic ensemble representations of conditionally folded IDRs. Significance StatementAlphaFold2 and other machine learning-based methods can accurately predict the structures of most proteins. However, nearly two-thirds of human proteins contain segments that are highly flexible and do not autonomously fold, otherwise known as intrinsically disordered regions (IDRs). In general, IDRs interconvert rapidly between a large number of different conformations, posing a significant problem for protein structure prediction methods that define one or a small number of stable conformations. Here, we found that AlphaFold2 can readily identify structures for a subset of IDRs that fold under certain conditions (conditional folding). We leverage AlphaFold2s predictions of conditionally folded IDRs to quantify the extent of conditional folding across the tree of life, and to rationalize disease-causing mutations in IDRs. Classifications: Biological Sciences; Biophysics and Computational Biology

biochemistry↗

bioRxiv2021

Discovering molecular features of intrinsically disordered regions by using evolution for contrastive learning

A major challenge to the characterization of intrinsically disordered regions (IDRs), which are widespread in the proteome, but relatively poorly understood, is the identification of molecular features that mediate functions of these regions, such as short motifs, amino acid repeats and physicochemical properties. Here, we introduce a proteome-scale feature discovery approach for IDRs. Our approach, which we call "reverse homology", exploits the principle that important functional features are conserved over evolution. We use this as a contrastive learning signal for deep learning: given a set of homologous IDRs, the neural network has to correctly choose a held-out homologue from another set of IDRs sampled randomly from the proteome. We pair reverse homology with a simple architecture and standard interpretation techniques, and show that the network learns conserved features of IDRs that can be interpreted as motifs, repeats, or bulk features like charge or amino acid propensities. We also show that our model can be used to produce visualizations of what residues and regions are most important to IDR function, generating hypotheses for uncharacterized IDRs. Our results suggest that feature discovery using unsupervised neural networks is a promising avenue to gain systematic insight into poorly understood protein sequences.

bioinformatics↗

bioRxiv2021

A functionally divergent intrinsically disordered region underlying the conservation of stochastic signaling

Stochastic signaling dynamics expand living cells information processing capabilities. An increasing number of studies report that regulators encode information in their pulsatile dynamics. The evolutionary mechanisms that lead to complex signaling dynamics remain uncharacterized, perhaps because key interactions of signaling proteins are encoded in intrinsically disordered regions (IDRs), whose evolution is difficult to analyze. Here we focused on the stochastic pulsing dynamics of Crz1, a transcription factor in fungi downstream of the widely conserved calcium signaling pathway. We find that Crz1 IDRs from anciently diverged fungi can all respond transiently to calcium stress; however, only Crz1 IDRs from the Saccharomyces clade support pulsatility, encode extra information, and rescue fitness, while the Crz1 IDRs from distantly related fungi do none of the three. On the other hand, we find that Crz1 pulsing is conserved in the distantly related fungi, consistent with the evolutionary model of stabilizing selection. Further, we show that a calcineurin docking site in a specific part of the IDRs appears to be sufficient for pulsing and show evidence for a beneficial increase in the relative calcineurin affinity of this docking site. We propose that evolutionary flexibility of functionally divergent IDRs underlies the conservation of stochastic signaling by stabilizing selection.

evolutionary biology↗

bioRxiv2020

Evolutionary dynamics of neutral phenotypes under DNA substitution models

We study the evolution of quantitative molecular traits in the absence of selection. Using Felsensteins 1981 DNA substitution model, we predict a linear restoring force on the mean of an additive phenotype. Remarkably, the mean dynamics are independent of the effect sizes and genotype and are similar to those predicted by the widely-used OU model for stabilizing selection and the house-of-cards model for phenotype evolution. We confirm the predictions empirically using additive molecular phenotypes calculated from ancestral reconstructions of putatively unconstrained DNA sequences in primate genomes. We predict and confirm empirically that the dynamics of the variance are more complicated than those predicted by the OU model, and show that our results for the restoring force of mutation hold even for nonadditive phenotypes, such as number of transcription factor binding sites, longest encoded peptide and folding propensity of the encoded peptide. Our results have implications for efforts to infer selection based on quantitative phenotype dynamics as well as to understand long-term trends in evolution of quantitative molecular traits.

evolutionary biology↗

bioRxiv2020

Self-Supervised Contrastive Learning of Protein Representations By Mutual Information Maximization

Pretrained embedding representations of biological sequences which capture meaningful properties can alleviate many problems associated with supervised learning in biology. We apply the principle of mutual information maximization between local and global information as a self-supervised pretraining signal for protein embeddings. To do so, we divide protein sequences into fixed size fragments, and train an autoregressive model to distinguish between subsequent fragments from the same protein and fragments from random proteins. Our model, CPCProt, achieves comparable performance to state-of-the-art self-supervised models for protein sequence embeddings on various downstream tasks, but reduces the number of parameters down to 2% to 10% of benchmarked models. Further, we explore how downstream assessment protocols affect embedding evaluation, and the effect of contrastive learning hyperparameters on empirical performance. We hope that these results will inform the development of contrastive learning methods in protein biology and other modalities.

bioinformatics↗

bioRxiv2020

Identifying molecular features that are associated with biological function of intrinsically disordered protein regions

Previously, we showed that intrinsically disordered regions in proteins (IDRs) contain multiple sequence-distributed molecular features that are conserved over evolution, despite little sequence similarity that can be detected in alignments (Zarin et al. 2019). Here, we aim to use these molecular features to make specific functional predictions for individual IDRs and identify the molecular features within them that are responsible for specific functions. We find that the predictable functions are diverse, identifying previously known associated molecular features, as well as features that were previously not known to be associated with these functions. We experimentally confirm that elevated isoelectric point and hydrophobicity, features that are positively associated with mitochondrial localization, are necessary for mitochondrial targeting function. Remarkably, increasing isoelectric point in a synthetic IDR restores weak mitochondrial targeting. We believe feature analysis represents a new systematic approach to understand how biological functions of IDRs are specified by their protein sequences.

cell biology↗