Long noncoding RNAs (lncRNAs) have emerged as central regulators of gene expression, genome organisation and cellular identity, yet the vast majority remain functionally uncharacterised. Their highly context-specific expression, incomplete annotation and diverse mechanisms of action have made them challenging to discover and study systematically. Addressing these challenges requires the development of computational and experimental methods specifically designed for lncRNA biology.
Here, we present an integrated framework combining bioinformatics, bulk and single-cell transcriptomics and CRISPR technologies to discover, annotate and functionally characterise lncRNAs in cancer. We first developed de novo transcriptome assembly pipelines integrating short-read and long-read RNA sequencing, enabling the discovery of tens of thousands of previously unannotated lncRNAs. Combined with deep, strand-specific and single-cell RNA sequencing, these approaches revealed that lncRNAs outperform protein-coding genes in defining breast cell populations and molecular subtypes while generating comprehensive atlases for downstream functional studies.
To systematically interrogate lncRNA function, we pioneered transcriptome-scale CRISPR-Cas13 pooled screens targeting more than 800 breast cancer-associated lncRNAs. These screens identified multiple regulators of tumour cell proliferation, including the tumour suppressor lncRNA KILR, which modulates DNA replication and repair by regulating RPA1 availability. We are currently expanding this platform to identify lncRNAs involved in therapeutic response and are currently extending these studies into three-dimensional tumour spheroids and in vivo models to better capture clinically relevant tumour biology.
Building on these advances, we are developing the next generation of functional genomics platforms for the noncoding genome. We are designing the first lncRNA-focused CROP-seq platform, coupling CRISPR-Cas13-mediated transcript silencing with single-cell transcriptomics to identify downstream target genes and regulatory pathways. In parallel, we are establishing pooled prime editing screens to functionally assess thousands of somatic mutations within lncRNAs, enabling direct investigation of how noncoding variants alter RNA function and drive cancer phenotypes. These experimental approaches are supported by dedicated computational pipelines for transcript discovery, guide RNA design, mutation prioritisation and large-scale screen analysis.
Together, these technologies establish a scalable framework for functional annotation of the noncoding genome. By integrating bioinformatics, CRISPR screening, single-cell genomics and genome editing, we aim to accelerate the discovery of functional lncRNAs, uncover noncoding drivers of cancer and lay the foundation for precision RNA medicine through the identification of novel biomarkers and RNA-based therapeutic targets.