Tommaso Biancalani - Sr Director & Head Of AI For Research Biology, AI/ML, Computational Sciences

Tommaso Biancalani

Sr Director & Head Of AI For Research Biology, AI/ML, Computational Sciences

Postdoc Mentor
"In the world of biology, machine learning is the compass that guides us through the wilderness of data."
Years at Genentech

I earned my PhD from the University of Manchester in 2013, specializing in mathematical physics with a research focus on stochastic behaviors of complex systems. Following that, I relocated to the United States for my postdoctoral training. During this period, I ventured into the realm of biology, initially at the Carl Woese Institute of Genomics and later at MIT, where I had the opportunity to establish my own experimental system (!). In 2018, I initiated a research group at the Broad Institute of MIT and Harvard, concentrating on the development of deep learning techniques for constructing the Human Cell Atlas. In 2021, I made the transition to Genentech and undertook the exciting task of building the BRAID organization from the ground up. In my other life, I pursue my passion for jazz guitar and indulge in the art of card magic. I also love cats.


BRAID (Biology Research | AI Development) operates as a division within gRED Computational Sciences, with a primary focus on leveraging machine learning to advance the field of biology. While our interests encompass a wide range of topics, our core mission revolves around target identification. More specifically, our department is actively engaged in ongoing research across the following applied domains:

  • Inferring gene functions using perturbation screens.
  • Exploring cell-cell communication through spatial transcriptomics.
  • Developing regulatory element designs through DNA sequence modeling.
  • Establishing foundational models for gene expression.

In addition to our applied research efforts, we maintain a dedicated theoretical section committed to advancing the theory and algorithms of machine learning.

Postdoctoral Mentor

The post-doctoral period represents a unique phase in a scientist's career, providing them with the opportunity to delve into their own research direction without the added responsibilities of leading a research group. Although these transitional years can be demanding, individuals can acquire the necessary skills to develop their ideas into fully-fledged research programs with the right mentorship and guidance. In my role as a dedicated mentor, I devote a substantial portion of my time to assisting our postdoctoral fellows in pinpointing pivotal questions that intersect the fields of machine learning and biology. Our department boasts a wealth of expertise and enjoys deep integration within the forefront of the biological community, making it an ideal environment for junior scientists to channel their theoretical skills into impactful research endeavors.

Featured Publication

Conformalized Deep Splines for Optimal and Efficient Prediction Sets.

arXiv preprint arXiv:2311.00774 (2023)

Nathaniel Diamant, Ehsan Hajiramezanali, Tommaso Biancalani, Gabriele Scalia.

Scalable querying of human cell atlases via a foundational model reveals commonalities across fibrosis-associated macrophages.

bioRxiv, 2023.07. 18.549537 (2023)

Heimberg, Kuo et al.

Improving Graph Generation by Restricting Graph Bandwidth.

ICML, 7939-7959 (2023)

Nathaniel Diamant, Alex M. Tseng, Kangway V. Chuang, Tommaso Biancalani, Gabriele Scalia.

Towards Understanding and Improving GFlowNet Training.

ICML: 30956–30975 (2023)

Max W. Shen, Emmanuel Bengio, Ehsan Hajiramezanali, Andreas Loukas, Kyunghyun Cho, Tommaso Biancalani.

NODAGS-Flow: Nonlinear Cyclic Causal Structure Learning.

AISTATS, 6371-6387 (2023)

Muralikrishnna G. Sethuraman, Romain Lopez, Rahul Mohan, Faramarz Fekri, Tommaso Biancalani, Jan-Christian Hütter.

Deep learning and alignment of spatially resolved single-cell transcriptomes with Tangram.

Nature Methods 18, 1352–1362 (2021)

Biancalani, Scalia et al.