Nicholas Ho

EMBER, a gene-centric foundation model, improves cross-context perturbation-response prediction and genetic-interaction ranking using contextualized gene embeddings.

This study evaluates foundation models for predicting cellular responses to perturbations, finding that some pretrained embeddings improve predictions while others do not.

This paper introduces HO_SCPLOWEIMDALLC_SCPLOW, a framework to analyze tokenizer design in scFMs, showing that specific tokenization choices improve generalization under distribution shift.

This paper introduces AIDO.DNA, a DNA language model that improves representation of DNA sequences across multiple tasks by scaling model size and training data.

The paper introduces AIDO.Cell, a series of dense Transformer models pre-trained on human single-cell RNA-seq data, achieving state-of-the-art results in clustering, classification, and perturbation modeling.