In bioinformatics, modeling the protein space to better predict function and structure has benefitted from Protein Language Models (PLMs). Their basis is the protein’s amino acid sequence and self-supervised learning. Ankh is a prime example of such a PLM. While there has been some recent work on integrating structure with a PLM to enhance predictive performance, to date there has been no work on integrating secondary structure rather than three-dimensional structure. Here we present TooT-PLM-P2S that begins with the Ankh model pre-trained on 45 million proteins using self-supervised learning. TooT-PLM-P2S builds upon the Ankh model by initially using its pre-trained encoder and decoder. It then undergoes an additional training phase with approximately 10,000 proteins and their corresponding secondary structures. This retraining process modifies the encoder and decoder, resulting in the creation of TooT-PLM-P2S. We then assess the impact of integrating secondary structure information into the Ankh model by comparing Ankh and TooT-PLM-P2S on eight downstream tasks including fluorescence and solubility prediction, sub-cellular localization, and membrane protein classification. For both Ankh and TooT-PLM-P2S the downstream tasks required task-specific training. Few of the results showed statistically significant differences. Ankh outperformed on three of the eight tasks, TooT-PLM-P2S did not outperform on any task for the primary metric. TooT-PLM-P2S did outperform for the precision metric for the task of discriminating membrane proteins from non-membrane proteins. This study requires future work with expanded datasets and refined integration methods.

This study presents TooT-PLM-ionCT, a holistic framework that exploits the capabilities of six diverse Protein Language Models (PLMs) - ProtBERT, ProtBERT-BFD, ESM-1b, ESM-2 (650M parameters), and ESM-2 (15B parameters) - for precise classification of integral membrane proteins, specifically ion channels (ICs) and ion transporters (ITs). As these proteins play a pivotal role in the regulation of ion movement across cellular membranes, they are integral to numerous biological processes and overall cellular vitality. To circumvent the costly and time-consuming nature of wet lab experiments, we harness the predictive prowess of PLMs, drawing parallels with techniques in natural language processing. Our strategy engages six classifiers, embracing both conventional methodologies and a deep learning model, to segregate ICs and ITs from other membrane proteins, as well as differentiate ICs from ITs. Furthermore, we delve into critical factors influencing our tasks, including the implications of dataset balancing, the effect of frozen versus fine-tuned PLM representations, and the potential variance between half and full precision floating-point computations. Our empirical results showcase superior performance in distinguishing ITs from other membrane proteins and differentiating ICs from ITs, while the task of discriminating ICs from other membrane proteins exhibits results commensurate with the current state-of-the-art.