Statistical Mechanical Model of the Giant Electrocaloric Effect in Ferroelectric Polymers.

The development of highly efficient cooling technologies has been identified as a key strategy to address the mitigation of global warming. Especially, electrocaloric materials have emerged as promising candidates for cooling applications, owing to their potential to provide high cooling capacity with low energy consumption. To advance the development of electrocaloric materials with a significant electrocaloric effect (ECE), a thorough understanding of the underlying mechanisms is required. Previous studies have estimated the maximum ECE temperature change by calculating the entropy change between two assumed states of a dipole model, assuming polarization saturation with a sufficiently large electric field. However, it is more relevant to assess the ECE under continuously changing electric fields as this is more reflective of real-world conditions. To this end, we establish a continuous transition between the complete disorder state and the polarization saturation state using the partition function to derive the entropy change. Our results demonstrate excellent agreement with experimental data, and our analysis of energy items within the partition function attributes the increase in the ECE entropy change with decreasing crystal size to interfacial effects. This statistical mechanical model reveals the in-depth ferroelectric polymers producing the ECE and offers significant potential for predicting the ECE in ferroelectric polymers and thus guides the design of high-performance ECE materials.

SESNet: sequence-structure feature-integrated deep learning method for data-efficient protein engineering.

Deep learning has been widely used for protein engineering. However, it is limited by the lack of sufficient experimental data to train an accurate model for predicting the functional fitness of high-order mutants. Here, we develop SESNet, a supervised deep-learning model to predict the fitness for protein mutants by leveraging both sequence and structure information, and exploiting attention mechanism. Our model integrates local evolutionary context from homologous sequences, the global evolutionary context encoding rich semantic from the universal protein sequence space and the structure information accounting for the microenvironment around each residue in a protein. We show that SESNet outperforms state-of-the-art models for predicting the sequence-function relationship on 26 deep mutational scanning datasets. More importantly, we propose a data augmentation strategy by leveraging the data from unsupervised models to pre-train our model. After that, our model can achieve strikingly high accuracy in prediction of the fitness of protein mutants, especially for the higher order variants (> 4 mutation sites), when finetuned by using only a small number of experimental mutation data (< 50). The strategy proposed is of great practical value as the required experimental effort, i.e., producing a few tens of experimental mutation data on a given protein, is generally affordable by an ordinary biochemical group and can be applied on almost any protein.