150813 research outputs found
Sort by
Incorporating Synthetic Accessibility in Drug Design: Predicting Reaction Yields of Suzuki Cross-Couplings by Leveraging AbbVie’s 15-Year Parallel Library Data Set
No full textDespite the increased use of computational tools to supplement medicinal chemists' expertise and intuition in drug design, predicting synthetic yields in medicinal chemistry endeavors remains an unsolved challenge. Existing design workflows could profoundly benefit from reaction yield prediction, as precious material waste could be reduced, and a greater number of relevant compounds could be delivered to advance the design, make, test, analyze (DMTA) cycle. In this work, we detail the evaluation of AbbVie's medicinal chemistry library data set to build machine learning models for the prediction of Suzuki coupling reaction yields. The combination of density functional theory (DFT)-derived features and Morgan fingerprints was identified to perform better than one-hot encoded baseline modeling, furnishing encouraging results. Overall, we observe modest generalization to unseen reactant structures within the 15-year retrospective library data set. Additionally, we compare predictions made by the model to those made by expert medicinal chemists, finding that the model can often predict both reaction success and reaction yields with greater accuracy. Finally, we demonstrate the application of this approach to suggest structurally and electronically similar building blocks to replace those predicted or observed to be unsuccessful prior to or after synthesis, respectively. The yield prediction model was used to select similar monomers predicted to have higher yields, resulting in greater synthesis efficiency of relevant drug-like molecules
Iron-sulfur clusters: the road to room temperature
No full textIron-sulfur proteins perform a wide variety of reactions central to the metabolisms of all living organisms. Foundational to their reaction chemistry are the rich electronic structures of their constituent Fe-S clusters, which differ in important ways from the active sites of mononuclear Fe enzymes. In this perspective, we summarize the essential electronic structure features that make Fe-S clusters unique, and point to the need for studies aimed at understanding the electronic basis for their reactivity under physiological conditions. Specifically, at ambient temperature, both the ground state and a large number of excited states are thermally populated, and thus a complete understanding of Fe-S cluster reactivity must take into account the properties, energies, and reactivity patterns of these excited states. We highlight prior research toward characterizing the low-energy excited states of Fe-S clusters that has established what is now a consensus model of these excited state manifolds and the bonding interactions that give rise to them. In particular, we discuss the low-energy alternate spin states and valence electron configurations that occur in Fe-S clusters of varying nuclearities, and finally suggest that there may be unrecognized functional roles for these states. Graphical abstrac
Crystalline Antibody‐Laden Alginate Particles: A Platform for Enabling High Concentration Subcutaneous Delivery of Antibodies
No full textSubcutaneous (SC) administration is a desired route for monoclonal antibodies (mAbs). However, formulating mAbs for small injection volumes at high concentrations with suitable stability and injectability is a significant challenge. Here, this work presents a platform technology that combines the stability of crystalline antibodies with injectability and tunability of soft hydrogel particles. Composite alginate hydrogel particles are generated via a gentle centrifugal encapsulation process which avoids use of chemical reactions or an external organic phase. Crystalline suspension of anti‐programmed cell death protein 1 (PD‐1) antibody (pembrolizumab) is utilized as a model therapeutic antibody. Crystalline forms of the mAb encapsuled in the hydrogel particles lead to stable, high concentration, and injectable formulations. Formulation concentrations as high as 315 mg mL−1 antibody are achieved with encapsulation efficiencies in the range of 89–97%, with no perceivable increase in the number of antibody aggregates. Bioanalytical studies confirm superior maintained quality of the antibody in comparison with formulation approaches involving organic phases and chemical reactions. This work illustrates tuning the alginate particles’ disintegration by using partially oxide alginates. Crystalline mAb‐laden particles are evaluated for their biocompatibility using cell‐based in vitro assays. Furthermore, the pharmacokinetics (PK) of the subcutaneously delivered human anti‐PD‐1 mAb in crystalline antibody‐laden alginate hydrogel particles in Wistar rats is evaluated
What’s for Dessert?
No full textDan and Tanya meet in a coffeehouse and decide to have dessert. Both are watching their calories, so they decide to share. They would like to find a dessert that they will both enjoy, and to do so quickly, with a minimum of negotiation or calculation. How should they choose
A functional approach for studying technological progress: Extension to energy technology
No full textThis paper extends a broad functional category approach for the study of technological capability progress recently developed and applied to information technology to a second key case—that of energy based technologies. The approach is applied to the same three functional operations—storage, transportation and transformation—that were used for information technology by first building a 100 plus year database for each of the three energy-based functional categories. In agreement with the results for information technology in the first paper, the energy technology results indicate that the functional approach offers a stable methodology for assessing longer time technological progress trends. Moreover, similar to what was found with information technology in the first study, the functional capability for energy technology shows continual—if not continuous—improvement that is best quantitatively described as exponential with respect to time. The absence of capability discontinuities—even with large technology displacement—and the lack of clear saturation effects are found with energy as it was with information. However, some key differences between energy and information technology are seen and these include:
*Lower rates of progress for energy technology over the entire period: 19–37% annually for Information Technology and 3–13% for Energy Technology.
*Substantial variability of progress rates is found within given functional categories for energy compared to relatively small variation within any one category for information technology. The strongest variation is found among capability progress among different energy types.
*More challenging data recovery and metric definition for energy as compared to information technology.
These findings are interpreted in terms of fundamental differences between energy and information including the losses and efficiency constraints on energy. We apply Whitney's insight that these fundamental differences lead to naturally modular information technology artifacts. The higher progress rates of information-based as opposed to energy-based technologies follows since decomposable systems can progress more rapidly due to the greater ease of independent as opposed to simultaneous development. In addition, the broad implications of our findings to studies of the relationships between technical and social change are briefly discussed
Integrating Machine Learning and Large Language Models to Advance Exploration of Electrochemical Reactions
No full textElectrochemical C−H oxidation reactions offer a sustainable route to functionalize hydrocarbons, yet identifying suitable substrates and optimizing synthesis remain challenging. Here, we report an integrated approach combining machine learning and large language models to streamline the exploration of electrochemical C−H oxidation reactions. Utilizing a batch rapid screening electrochemical platform, we evaluated a wide range of reactions, initially classifying substrates by their reactivity, while LLMs text‐mined literature data to augment the training set. The resulting ML models for reactivity prediction achieved high accuracy (>90 %) and enabled virtual screening of a large set of commercially available molecules. To optimize reaction conditions for selected substrates, LLMs were prompted to generate code that iteratively improved yields. This human‐AI collaboration proved effective, efficiently identifying high‐yield conditions for 8 drug‐like substances or intermediates. Notably, we benchmarked the accuracy and reliability of 12 different LLMs–including LLaMA series, Claude series, OpenAI o1, and GPT‐4‐on code generation and function calling related to ML based on natural language prompts given by chemists to showcase potentials for accelerating research across four diverse tasks. In addition, we collected an experimental benchmark dataset comprising 1071 reaction conditions and yields for electrochemical C−H oxidation reactions
Report to the President for year ended June 30, 2025, Vice Provost for the Arts
No full textThis report contains the following sections: Artfinity arts festival, Arts Initiatives, Center for Art, Science & Technology (CAST), the List Visual Arts Center, the MIT Museum, Administrative Initiatives, Finances and Funding, and Personnel
From Parking to Parcels: The Potential for Microhubs in New York City’s Parking Garages
No full textThis thesis employs a site planning and policy perspective to explore how parking garages can serve as last-mile microhubs for e-commerce package deliveries in New York City. During the COVID-19 pandemic, deliveries accelerated, prompting a proliferation of “last-mile facilities,” the destination where parcels go just prior to final delivery. This surge of activity has prompted residents to raise complaints about trucks and vans driving through their neighborhoods and blocking streets or sidewalks when unloading their goods. In response, New York City government has been forced to think more proactively about the freight supply chain and its impact on the urban environment. New York and other cities have begun experimenting with the use of microhubs. Microhubs are small spaces in which packages are unloaded from vans and trucks onto smaller, more sustainable modes such as cargo bikes and handcarts. A commonly identified but understudied location for microhubs is the parking garage. London stands out as a city with this form of hub. This thesis employs three primary research methods—site observations, interviews, and case studies—to argue that parking garages could provide a solution to better utilize dense urban space in dense cities and improve quality of life for residents by reducing the negative impacts of existing last-mile warehouses and delivery vehicles, all while requiring minimal funding. This is shown through an analysis of existing microhub sites in London and how they relate to their urban surroundings. These findings are then applied to two distinct contexts and garage designs in New York City. Finally, the thesis offers site planning criteria that connect land use policy to the design of the facilities and the surrounding public realm through the concept of “planning at the interface.”M.C.P
Paleo-aridity records investigated through drip water chemistry in Lehman Caves, Nevada
No full textIn the Great Basin of the southwest United States, climate change is predicted to cause increased precipitation variability, making the future climate of the region uncertain. The paleoclimate record has direct examples of dramatic changes in water availability in this area, allowing for a comparison of precipitation changes and responses for the Great Basin. In Lehman Cave, Nevada, ten drips above actively-forming stalagmites were sampled monthly. Glass growth plates were also placed above three actively-forming stalagmites, allowing for the collection of new calcite growths. This project analyzed the samples for Mg/Ca, Sr/Ca, and U/Ca ratios to provide a comparison of the composition of calcite and the drip waters from which they precipitate. This will improve our understanding of the paleo-aridity of the Great Basin region, as well as provide useful context for the changing precipitation patterns expected with modern climate change.S.B