Microstructural Evolution in a 6060 Extrudable Al-Alloy: Integrated Modeling and Experimental Validation.

Desirable properties including strength, ductility and extrudability of 6060 Al-alloys are highly dependent on processing to control the development of microstructural features. In this study, the process chain of an extrudable 6060 Al-alloy was modeled in an Integrated Computational Materials Engineering framework and validated experimentally via quantitative SEM-EDX and TEM. All critical processing stages were considered including casting, homogenization heating and holding, extrusion cooling and two-stage aging. Segregation and intermetallics formation were accurately predicted and experimentally verified in the as-cast condition. Diffusion simulations predicted the dissolution of intermetallics and completion of ß-AlFeSi to α-AlFeSi transformation during homogenization, in excellent agreement with quantitative SEM-EDX characterization. Precipitation simulations predicted the development of a ß″ strengthening dispersion during extrusion cooling and aging. Needle-shaped ß″ precipitates were observed and analyzed with quantitative high-resolution TEM, validating predictions. Ensuing precipitation strengthening was modeled in terms of aging time, presenting good agreement with yield strength measurements. Precipitate-Free Zones and coarse, metastable ß-type particles on dispersoids and grain boundaries were investigated. The proposed integrated modeling and characterization approach considers all critical processing stages and could be used to optimize processing of extrudable 6xxx Al-alloys, providing insight to mechanisms controlling microstructural evolution and resulting properties.

cdsBERT - Extending Protein Language Models with Codon Awareness.

Recent advancements in Protein Language Models (pLMs) have enabled high-throughput analysis of proteins through primary sequence alone. At the same time, newfound evidence illustrates that codon usage bias is remarkably predictive and can even change the final structure of a protein. Here, we explore these findings by extending the traditional vocabulary of pLMs from amino acids to codons to encapsulate more information inside CoDing Sequences (CDS). We build upon traditional transfer learning techniques with a novel pipeline of token embedding matrix seeding, masked language modeling, and student-teacher knowledge distillation, called MELD. This transformed the pretrained ProtBERT into cdsBERT; a pLM with a codon vocabulary trained on a massive corpus of CDS. Interestingly, cdsBERT variants produced a highly biochemically relevant latent space, outperforming their amino acid-based counterparts on enzyme commission number prediction. Further analysis revealed that synonymous codon token embeddings moved distinctly in the embedding space, showcasing unique additions of information across broad phylogeny inside these traditionally "silent" mutations. This embedding movement correlated significantly with average usage bias across phylogeny. Future fine-tuned organism-specific codon pLMs may potentially have a more significant increase in codon usage fidelity. This work enables an exciting potential in using the codon vocabulary to improve current state-of-the-art structure and function prediction that necessitates the creation of a codon pLM foundation model alongside the addition of high-quality CDS to large-scale protein databases.