CITIC

International researchers highlight breakthroughs in microbiome and natural language processing at the TIC Talk Breakfast Series

30/09/2025 - CITIC

CITIC hosted a new session of its TIC Talk Breakfast on Tuesday, September 30, where international researchers shared cutting-edge innovations combining data science, biomedicine, and language processing — illustrating how technology can transform both research and clinical practice.

Ayesha Wasim, a doctoral researcher at CITIC under the MCSA COFUND program, presented her work on algorithms and data analysis applied to the microbiome in colorectal cancer. Her research focuses on developing automated and reproducible processes capable of identifying consistent microbial patterns across multiple hospitals — a crucial step toward validating reliable clinical biomarkers. Among her findings, Wasim highlighted several bacterial species whose presence may influence cancer development through mechanisms related to inflammation, metabolism, and the immune response. Looking ahead, her next challenge will be to distinguish pathogenic strains from commensals and link microbial genes to key metabolic and immunological pathways.

Muhammad Imran, also a researcher funded by the Marie Curie program, introduced an innovative natural language processing (NLP) approach that combines precision with high speed. His method reframes syntactic parsing as a sequence-labeling task, enabling the processing of up to 1,000 sentences per second without sacrificing accuracy. This advancement not only enhances sentiment analysis but also significantly improves entity recognition in biomedical texts, outperforming several traditional NLP systems in benchmark tests.

Both speakers underscored a central message: efficiency and reproducibility are essential to translating scientific progress into real-world applications, whether in medicine or technology.