Journal
JOURNAL OF APPLIED PHYSICS
Volume 134, Issue 8, Pages -Publisher
AIP Publishing
DOI: 10.1063/5.0157367
Keywords
-
Categories
Ask authors/readers for more resources
We introduce a flexible deep learning strategy based on a language model to solve complex forward and inverse problems in protein modeling. The model combines transformer and graph convolutional architectures to create a generative pretrained model. It can predict protein structural properties, solubility, and sequence tasks. By training it on inverse tasks, the model can also be used to design proteins with desired properties. The model is prompt-based and can be adapted for various downstream tasks. The study demonstrates the importance of adding additional tasks to improve model performance. The model has been validated through case studies and has shown potential in designing structural and antimicrobial biomaterials.
We report a flexible language-model-based deep learning strategy, applied here to solve complex forward and inverse problems in protein modeling, based on an attention neural network that integrates transformer and graph convolutional architectures in a causal multi-headed graph mechanism, to realize a generative pretrained model. The model is applied to predict the secondary structure content (per-residue level and overall content), protein solubility, and sequencing tasks. Further trained on inverse tasks, the model is rendered capable of designing proteins with these properties as target features. The model is formulated as a general framework, completely prompt-based, and can be adapted for a variety of downstream tasks. We find that adding additional tasks yields emergent synergies that the model exploits in improving overall performance, beyond what would be possible by training a model on each dataset alone. Case studies are presented to validate the method, yielding protein designs specifically focused on structural materials, but also exploring the applicability in the design of soluble, antimicrobial biomaterials. While our model is trained to ultimately perform eight distinct tasks, with available datasets, it can be extended to solve additional problems. In a broader sense, this study illustrates a form of multiscale modeling that relates a set of ultimate building blocks (here, byte-level utf8 characters that define the nature of the physical system at hand) to complex output. This materiomic scheme captures complex emergent relationships between universal building block and resulting properties, via a synergizing learning capacity, to express a set of potentialities embedded in the knowledge used in training via the interplay of universality and diversity. Significance statement: Predicting the properties of materials based on a flexible description of their structure, environment, or process, is a long-standing challenge in multiscale modeling. Our MaterioFormer language model, trained to solve forward and inverse tasks, incorporates a deep learning capacity through attention and graph strategies to yield a multimodal approach to model and design materials. Since our model is prompt-based and information is encoded consistently via byte-level utf8 tokenization, it can process diverse modalities of information, such as sequence data, description of tasks, and numbers, and offers a flexible workflow that integrates human intelligence and artificial intelligence. Autoregressive training, using pre-training against a large unlabeled dataset, allows for straightforward adjustment of specific objectives. (c) 2023 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available