About Us

We are an interdisciplinary research group at Academia Sinica, co-led by I-Ping Tu (Institute of Statistical Science) and Wei-Hau Chang (Institute of Chemistry). Our work bridges statistical innovation and structural biology, with a focus on the analysis of cryo-electron microscopy (cryo-EM) data.

Since 2014, we have developed a series of robust and efficient statistical methods to address core challenges in cryo-EM, including dimension reduction, image clustering, and 3D structure classification from noisy 2D particle images. These advances earned international recognition, highlighted by the 2020 ICCM Best Paper Award.

Beyond methodology, we have provided critical insights into fundamental issues such as the “Einstein from Noise” phenomenon, where we quantified and corrected bias in cryo-EM reconstructions. Our tools—most notably the RE2DC classification package, integrated into the ASCEP processing platform—have been applied to real-world structural biology problems, including uncovering the mechanism of filament dynamics in homologous recombination from low-resolution cryo-EM data.

Building on this foundation, our group is now advancing methods to move beyond coordinate fitting toward extracting chemically and electrostatically meaningful atomic features directly from cryo-EM maps, aiming to transform cryo-EM into a platform for quantitative chemical and mechanistic insight.