arXiv subjects

Alan M. Moses

Publications and source records attributed to Alan M. Moses.

7 recordsLinked to original sources

arXiv2026

C3P: Contrastive promoter-protein pretraining yields representations capturing bacterial gene regulation

Despite the increasing scale of genome language models (gLMs), their ability to decode the function of regulatory sequences remains unclear. gLM pretraining relies on sequence reconstruction, which may struggle due to the noisy, rapidly evolving nature of regulatory DNA. Self-supervised contrastive approaches provide a promising alternative. Inspired by language-image architectures like CLIP, we introduce contrastive promoter-protein pretraining (C3P). By learning to align promoters to their corresponding proteins, we leverage the rich representations of proteins learned by protein language models as supervisory signal for the learning of promoter representations. After training on 88 million bacterial promoter-protein pairs, we evaluate the predictive power of C3P-learned promoter representations for inference of curated regulatory annotations, finding multi-fold improvement over leading gLMs. We also introduce zero-shot co-regulated gene retrieval, the ability to find co-regulated genes in a genome using no experimental data. We find that compared to a randomly initialized baseline, C3P training consistently provides significant zero-shot performance gains, unlike gLMs. Scaling analysis reveals the potential for further improvement as well as the efficiency of C3P, which achieved strong performance at a fraction of the training cost of leading gLMs. In addition to demonstrating that C3P training is effective for learning representations of bacterial regulatory sequences, our strong zero-shot co-regulated gene retrieval performance suggests the possibility of decoding gene regulation for millions of bacteria from their genomes alone.

q-bio.GN↗

arXiv2024

An RNA condensate model for the origin of life

The RNA World hypothesis predicts that self-replicating RNAs evolved before DNA genomes and coded proteins. Despite widespread support for the RNA World, self-replicating RNAs have yet to be identified in a natural context, leaving a key 'missing link' for this explanation of the origin of life. Inspired by recent work showing that condensates of charged polymers are capable of catalyzing chemical reactions, we consider a catalytic RNA condensate as a candidate for the self-replicating RNA. Specifically, we propose that short, low-complexity RNA polymers formed catalytic condensates capable of templated RNA polymerization. Because the condensate properties depend on the RNA sequences, RNAs that formed condensates with improved polymerization and demixing capacity would be amplified, leading to a 'condensate chain reaction' and evolution by natural selection. Many of the needed properties of this self-replicating RNA condensate have been realized experimentally in recent studies and our predictions could be tested with current experimental and theoretical tools. Our theory addresses central problems in the origins of life: (i) the origin of compartmentalization, (ii) the error threshold for the accuracy of templated replication, (iii) the free energy cost of maintaining an information-rich population of replicating RNA polymers. Furthermore, we note that the extant nucleolus appears to satisfy many of the requirements of an evolutionary relic for the model we propose. More generally, we suggest that future work on the origin of life would benefit from condensate-centric biophysical models of RNA evolution.

q-bio.BM↗

arXiv2021

CytoImageNet: A large-scale pretraining dataset for bioimage transfer learning

Motivation: In recent years, image-based biological assays have steadily become high-throughput, sparking a need for fast automated methods to extract biologically-meaningful information from hundreds of thousands of images. Taking inspiration from the success of ImageNet, we curate CytoImageNet, a large-scale dataset of openly-sourced and weakly-labeled microscopy images (890K images, 894 classes). Pretraining on CytoImageNet yields features that are competitive to ImageNet features on downstream microscopy classification tasks. We show evidence that CytoImageNet features capture information not available in ImageNet-trained features. The dataset is made available at https://www.kaggle.com/stanleyhua/cytoimagenet.

cs.CV↗

arXiv2021

Random Embeddings and Linear Regression can Predict Protein Function

Large self-supervised models pretrained on millions of protein sequences have recently gained popularity in generating embeddings of protein sequences for protein function prediction. However, the absence of random baselines makes it difficult to conclude whether pretraining has learned useful information for protein function prediction. Here we show that one-hot encoding and random embeddings, both of which do not require any pretraining, are strong baselines for protein function prediction across 14 diverse sequence-to-function tasks.

q-bio.BM↗

arXiv2020

W-Cell-Net: Multi-frame Interpolation of Cellular Microscopy Videos

Deep Neural Networks are increasingly used in video frame interpolation tasks such as frame rate changes as well as generating fake face videos. Our project aims to apply recent advances in Deep video interpolation to increase the temporal resolution of fluorescent microscopy time-lapse movies. To our knowledge, there is no previous work that uses Convolutional Neural Networks (CNN) to generate frames between two consecutive microscopy images. We propose a fully convolutional autoencoder network that takes as input two images and generates upto seven intermediate images. Our architecture has two encoders each with a skip connection to a single decoder. We evaluate the performance of several variants of our model that differ in network architecture and loss function. Our best model out-performs state of the art video frame interpolation algorithms. We also show qualitative and quantitative comparisons with state-of-the-art video frame interpolation algorithms. We believe deep video interpolation represents a new approach to improve the time-resolution of fluorescent microscopy.

eess.IV↗

arXiv2020

The Cells Out of Sample (COOS) dataset and benchmarks for measuring out-of-sample generalization of image classifiers

Understanding if classifiers generalize to out-of-sample datasets is a central problem in machine learning. Microscopy images provide a standardized way to measure the generalization capacity of image classifiers, as we can image the same classes of objects under increasingly divergent, but controlled factors of variation. We created a public dataset of 132,209 images of mouse cells, COOS-7 (Cells Out Of Sample 7-Class). COOS-7 provides a classification setting where four test datasets have increasing degrees of covariate shift: some images are random subsets of the training data, while others are from experiments reproduced months later and imaged by different instruments. We benchmarked a range of classification models using different representations, including transferred neural network features, end-to-end classification with a supervised deep CNN, and features from a self-supervised CNN. While most classifiers perform well on test datasets similar to the training dataset, all classifiers failed to generalize their performance to datasets with greater covariate shifts. These baselines highlight the challenges of covariate shifts in image data, and establish metrics for improving the generalization capacity of image classifiers.

cs.LG↗

arXiv2010

Towards a theoretical understanding of false positives in DNA motif finding

Detection of false-positive motifs is one of the main causes of low performance in motif finding methods. It is generally assumed that false-positives are mostly due to algorithmic weakness of motif-finders. Here, however, we derive the theoretical dependence of false positives on dataset size and find that false positives can arise as a result of large dataset size, irrespective of the algorithm used. Interestingly, the false-positive strength depends more on the number of sequences in the dataset than it does on the sequence length. As expected, false-positives can be reduced by decreasing the sequence length or by adding more sequences to the dataset. The dependence on number of sequences, however, diminishes and reaches a plateau after which adding more sequences to the dataset does not reduce the false-positive rate significantly. Based on the theoretical results presented here, we provide a number of intuitive rules of thumb that may be used to enhance motif-finding results in practice.

q-bio.GN↗