Antibodies have the capacity to bind a diverse set of antigens, and they have become critical therapeutics and diagnostic molecules. The binding of antibodies is facilitated by a set of six hypervariable loops that are diversified through genetic recombination and mutation. Even with recent advances, accurate structural prediction of these loops remains a challenge. Here, we present IgFold, a fast deep learning method for antibody structure prediction. IgFold consists of a pre-trained language model trained on 558M natural antibody sequences followed by graph networks that directly predict backbone atom coordinates. IgFold predicts structures of similar or better quality than alternative methods (including AlphaFold) in significantly less time (under one minute). Accurate structure prediction on this timescale makes possible avenues of investigation that were previously infeasible. As a demonstration of IgFold’s capabilities, we predicted structures for 105K paired antibody sequences, expanding the observed antibody structural space by over 40 fold.

Antibodies have the capacity to bind a diverse set of antigens, and they have become critical therapeutics and diagnostic molecules. The binding of antibodies is facilitated by a set of six hypervariable loops that are diversified through genetic recombination and mutation. Even with recent advances, accurate structural prediction of these loops remains a challenge. Here, we present IgFold, a fast deep learning method for antibody structure prediction. IgFold consists of a pre-trained language model trained on 558 million natural antibody sequences followed by graph networks that directly predict backbone atom coordinates. IgFold predicts structures of similar or better quality than alternative methods (including AlphaFold) in significantly less time (under 25 s). Accurate structure prediction on this timescale makes possible avenues of investigation that were previously infeasible. As a demonstration of IgFold’s capabilities, we predicted structures for 1.4 million paired antibody sequences, providing structural insights to 500-fold more antibodies than have experimentally determined structures.

The diffusion coefficient of methane in water plays an important role in the formation and dissociation of methane hydrate. However, most of the previous studies on the diffusion coefficient of methane in brine are performed at room temperature and low pressures, which is quite different from the formation condition of methane hydrate. In this study, we measure the diffusion coefficient of methane in pure water and brine in capillary tube at 10.3 MPa and temperature ranging from 283.15 to 308.15 K. We use the Raman spectrum to measure the ratio of C-H bound signal of methane to the O-H bound signal of water, to estimate the concentration of methane dissolves in water/brine. The Raman spectrum is collected at different time and different positions away from the liquid-gas interface. Diffusion coefficient is determined by fitting the experimental data with the concentration profiles solved from Fick's second law and semi-infinity boundary condition. By this method, we can evaluate the diffusion coefficient at different temperatures or salinities. The diffusion coefficient of methane in water/brine increases as the temperature increases. The diffusion coefficient of methane in brine is lower than that in pure water. Molecular dynamics (MD) simulation is also performed in this study to calculate the diffusion coefficient of methane in water/brine. The MD results can successfully predict the tendency of temperature effect and adding electrolyte.

We present the results for CAPRI Round 54, the 5th joint CASP-CAPRI protein assembly prediction challenge. The Round offered 37 targets, including 14 homo-dimers, 3 homo-trimers, 13 hetero-dimers including 3 antibody-antigen complexes, and 7 large assemblies. On average ~70 CASP and CAPRI predictor groups, including more than 20 automatics servers, submitted models for each target. A total of 21941 models submitted by these groups and by 15 CAPRI scorer groups were evaluated using the CAPRI model quality measures and the DockQ score consolidating these measures. The prediction performance was quantified by a weighted score based on the number of models of acceptable quality or higher submitted by each group among their 5 best models. Results show substantial progress achieved across a significant fraction of the 60+ participating groups. High-quality models were produced for about 40% for the targets compared to 8% two years earlier, a remarkable improvement resulting from the wide use of the AlphaFold2 and AlphaFold-Multimer software. Creative use was made of the deep learning inference engines affording the sampling of a much larger number of models and enriching the multiple sequence alignments with sequences from various sources. Wide use was also made of the AlphaFold confidence metrics to rank models, permitting top performing groups to exceed the results of the public AlphaFold-Multimer version used as a yard stick. This notwithstanding, performance remained poor for complexes with antibodies and nanobodies, where evolutionary relationships between the binding partners are lacking, and for complexes featuring conformational flexibility, clearly indicating that the prediction of protein complexes remains a challenging problem.

Guest migration in clathrate hydrates is a slow but important process for reaching thermodynamic equilibrium. The transport of guest molecules in a hydrate lattice is considered as a series of hopping events from an occupied cage to an empty neighboring cage facilitated by water vacancies and without significant lattice restructuring in the bulk. In this work, we developed an analytical model for determining the equilibrium distribution and the diffusivity of gas molecules in the cages of sI clathrate hydrate based on their hopping rate. Furthermore, kinetic Monte Carlo simulations were performed to verify the analytical model. The equilibrium occupancies, transport (Fickian) and jump (Maxwell–Stefan) diffusion coefficients, and the thermodynamic correction factors determined from the analytical model are in excellent agreement with the simulation results. Using the hopping rate constants determined based on transition path sampling calculations, we obtain the methane transport-diffusion coefficient at 275 K to be 5.06 × 10–14 m2/s, which is in good agreement with the recent experimental measurements (4.00 × 10–14 m2/s at 275 K). In addition, the analytical model is useful for identifying the predominant cage hopping pathways for equilibrium and transport properties.