Connor Coley (MIT)
Computer-aided synthesis planning has been pursued since the 1960s, beginning with expert systems that applied hand-encoded reaction rules to explore retrosynthetic pathways. Modern data-driven approaches learn transformations directly from large reaction databases, yet they struggle to propose correct disconnections for the complex targets increasingly common in small-molecule drug discovery. In this talk, I will describe how "higher-level" retrosynthetic strategies, which defer tactical decisions to a later stage, can yield shorter routes and solve more targets. I will then turn to the forward problem of predicting reaction outcomes, which is central to assessing the feasibility of proposed routes. I will contrast broadly trained models that generalize poorly with reaction-type-specific models built on purpose-generated data, physics-based methods accelerated by machine-learned interatomic potentials, and mechanistic deep learning models that enforce conservation of mass and electrons. Finally, I will discuss how combining imperfect reaction predictors with mass spectrometry-based structure elucidation may offer a path toward automated minor product and impurity identification.
Mingji Dai (Emory University)
This talk will focus on our recent efforts in function and efficiency-driven total synthesis of medicinally important natural products. The target molecules include macrolides, alkaloids, and polycyclic diterpenoids. Novel and enabling synthetic strategies and methodologies including skeletal editing, radical retrosynthesis, and catalytic carbonylation toward these target molecules will be highlighted. Biological evaluation and target identification of the selected natural products and their analogs will be discussed as well. These biological endeavors have led us to the exciting territories of targeting the previously undruggable disease proteins for novel therapeutic development. Overall, this talk will emphasize how we use natural product total synthesis to achieve chemistry innovation and biological discovery.
Betsy Parkinson (Purdue University)
Natural products (NPs) from the soil dwelling Actinomycetota are a bountiful source of bioactive molecules, including medicines, agricultural products, and chemical tools to study biological processes. Additionally, the biosynthetic enzymes that produce them perform unique chemistries and are excellent starting points for biocatalysts. Unfortunately, many biosynthetic gene clusters (BGCs, clusters of genes that encode for the enzymes that produce NPs) are silent. Accessing NPs from these silent BGCs is quite challenging, thus slowing the discovery of novel bioactive NPs that can serve as leads for medicines and agricultural products. Herein, we are exploring two different methods to access NPs from cryptic BGCs: 1) bioinformatics inspired direct chemical synthesis and 2) BGC induction using quorum sensing molecules. Additionally, we are discovering and utilizing biosynthetic enzymes to access otherwise challenging to access products, with a particular focus on cyclic peptides. Overall, these approaches enable us to access molecules that are otherwise inaccessible, thus helping to expand the medicinal and agricultural pipelines.
Kevin Brown (Indiana University)
Boron chemistry is an indispensable tool in modern organic synthesis, leveraging boron's unique electronic and structural properties. Established applications, such as the hydroboration reaction and the Suzuki-Miyaura cross-coupling, highlight its broad utility. This seminar will delve into novel uses of boron that enable the rapid synthesis of complex molecules with broad applications, including the development of alkene carboboration reactions for synthesizing essential building blocks in drug discovery, boron's role in directing cycloaddition reactions and C-H functionalization, and the application of boron in the modification of peptides. The discussion will cover mechanistic details, practical applications, and examples of industry collaboration, showcasing how these advancements contribute to modern chemical synthesis.
Quentin Michaudel (Texas A&M University)
Our research develops catalytic strategies for the precise synthesis of molecules and polymers with programmable structure and function, targeting applications from healthcare to sustainable materials and information technologies. This presentation will first describe our efforts centered on Sulfur(VI) Fluoride Exchange (SuFEx) chemistry as an enabling platform for editable click reactions, in which otherwise inert linkers serve as versatile synthetic handles for late-stage diversification. Specifically, we developed a photocatalytic deaminative C–C cross-coupling platform that converts primary amines into 1,2-dialkyldiazenes, providing modular access to diverse carbon frameworks. We will then discuss stereoselective olefin metathesis as a strategy to precisely control alkene geometry and sequence in conjugated polymers. Combining stereochemical control with wavelength- and dose-dependent photoisomerization enables multidimensional encoding of molecular states, which can be spatially programmed by digital light projection. Fluorescence provides a direct optical readout of these encoded states, illustrating how precise control of polymer stereochemistry can translate into programmable optical function.
Tianning Diao (New York University)
Nickel-catalyzed cross-coupling reactions have emerged as sustainable and cost-effective synthetic methods. While nickel operates via classical closed-shell pathways to catalyze biaryl coupling reactions, its ability to mediate radical pathways has significantly broadened the scope of cross-coupling to include a diverse range of alkyl motifs. Building on mechanistic insights, we developed the ProPhos ligand, which enhances the reactivity of nickel-catalyzed Suzuki-Miyaura couplings. This advancement paves the way for applying nickel catalysts in pharmaceutical process synthesis of heteroaromatic molecules. In parallel, our studies on nickel-mediated radical pathways elucidated the mechanistic details of radical initiation, capture, and bond formation. Leveraging this insight, we developed deoxygenative cross-coupling reactions that convert the hydroxyl group of serine and homoserine into a wide range of noncanonical functionalities, expanding the chemical space of noncanonical peptides.