My researches address the development of a practical approach and a computational platform for environmental risk assessment of Engineered Nanomaterials (ENMs), which increasing use in modern industrial products and processes has raised public concern regarding their potential release and damage to the environment. Development of such a platform requires the researches for High Throughput Screening (HTS) data processing methods (e.g., outlier removal, normalization, hit-identification), data mining for ENMs bioactivity data (e.g., clustering analysis for similar ENMs and exploring relationships among ENMs bioactivities), and decision support approach for ENMs environmental risk assessment.
I am also participating in nanoinformatics initiatives and developing tools/resources for integration and management of heterogeneous information, defining ENMs ontology, and modeling based knowledge extraction. In particular, I am interested in the development of (Quantitative)-Structure-Activity Relationships ((Q)SARs)) that can predict ENMs bioactivity from their structural and physicochemical properties and assist in the understanding of the mechanisms governing the behavior of ENMs in biological/ecological systems.
In addition to ENMs environmental risk assessment, my research interests also include general machine learning/data mining topics, such as development of supervised and unsupervised feature selection methods (to identify suitable features/descriptors for machine learning/data mining model development) and pattern recognition method for biological images.