Johns Hopkins researchers have introduced a revolutionary artificial intelligence tool known as Splam that is set to transform the way scientists study genetics. This new technology allows for precise identification of splicing sites in genes, a crucial aspect of understanding how genetic material is interpreted by cells and how mutations can lead to diseases.
Splicing is an essential biological process in which cells cut out non-essential parts of the gene, called introns, and join together the important sections, known as exons, to create functional proteins. This process ensures that genes are accurately translated into the proteins needed for the body’s functions. However, the ability to accurately pinpoint where splicing occurs has remained a challenge for researchers, until now.
Splam, developed by a team led by doctoral student Kuan-Hao Chao from the Whiting School of Engineering at Johns Hopkins, offers a breakthrough solution. Unlike older tools, Splam requires a much smaller DNA sequence to predict splice sites. It analyzes a compact 800-nucleotide sequence, compared to the larger, more cumbersome sequences used by previous methods like SpliceAI. This efficiency not only makes Splam more feasible for research but also more biologically accurate, as it mirrors the actual mechanisms that take place in cells.
The new tool has shown great potential in improving the accuracy of RNA sequencing, a technique used to measure gene expression and its variation in different conditions. This is particularly valuable in fields such as cancer research, where scientists compare gene activity in healthy cells versus cancerous ones to identify potential therapeutic targets. Splam can also enhance the process of genome annotation, which is critical for understanding how genes contribute to various biological functions.
To ensure its effectiveness beyond human DNA, the researchers also tested Splam on genetic material from other species, including a chimpanzee, a mouse, and a plant. The results were promising, demonstrating that Splam could accurately detect splice sites even in the more distantly related genomes, confirming its ability to generalize across different organisms.
With its potential to reduce errors and noise in transcriptome assembly, Splam could play a pivotal role in genomic studies and genetic testing, such as those used by companies like 23andMe. The team is now focused on further refining the tool and integrating it into existing RNA sequencing workflows to improve the study of gene function and disease.
Ultimately, Splam represents a significant leap forward in computational biology, offering a more effective, precise, and scalable approach to understanding the intricate processes of gene splicing. As this technology continues to evolve, it has the potential to revolutionize the way scientists study genetics and its applications in health and disease.
Subscribe to our newsletter!
