Top 10 place for Michele Tinti in machine learning competition
Published on 19 October 2022
Dr Michele Tinti recently participated in a biomedicine crowdsourced challenge instigated by DREAM Challenges. Michele was placed 10th out of almost 300 registered participants and 26 finalist teams.
The DREAM Challenges is a community-driven organization whose mission is to promote open, reproducible and collaborative scientific research in biomedicine. The outputs of the Challenges are publications, open-source code, data, documentation, and benchmarks about the methods used to address questions in healthcare and the life sciences. In the challenge Michele participated in, competitors were given expression measurements of millions of randomly generated promoter sequences to train machine learning models that predict gene expression from sequences.
Michele, a Senior Bioinformatician in the Division of Biological Chemistry and Drug Discovery explained why this competition was needed, “Despite recent progress in the field, there is room for improvement in deep learning models used for DNA sequences. Decoding how gene and regulatory DNA sequences work is critical to understanding the biology of organisms and diseases. For this reason, developing a good deep-learning framework will help the scientific community pinpoint regulatory sequences and understand how they work.
Michele continued, “It has been a great experience to compete with several laboratories around the world to solve this bioinformatics task. I’m looking forward to presenting my findings at the International Society for Computational Biology annual meeting RSGDREAM 2022 next month. A big thanks goes to Dr. Susan Wyllie, Prof. David Horn and Prof. Mike Ferguson for the support I received. It will be amazing to apply my findings and those of other participants to the field of neglected diseases. The models developed can help to understand the transcription regulation of neglected tropical disease agents and could also shed light on the working function of DNA/RNA regions important for post-transcription regulation.”