Alexandra Popa

SPEAK identifies spatial domains in spatial transcriptomics data using a large language model and expert-aligned prompts, showing improved accuracy and interpretability.

This study creates a biliary tract cancer cell line atlas using multi-omics data to identify molecular subtypes and therapeutic targets.

The paper identifies genetic vulnerabilities linked to transcription factor activities in cancer cell lines using CRISPR knockout data and validates targets associated with TEAD1.

This paper introduces ARTi, a method using synthetic shRNAs to achieve precise and controllable RNAi for gene suppression.

The paper identifies novel paralog gene dependencies in cancer, including sex chromosome-encoded gene pairs, suggesting therapeutic strategies for tumors with Y-chromosome loss.

The paper quantifies the fraction of alternative splicing events that are errors by analyzing transcriptomes of Paramecium tetraurelia and humans, finding that 92%-98% of splice variants in medium-expressed genes are errors.