Recently, diffusion- and flow-based generative models of protein structures have emerged as a powerful tool for de novo protein design. Here, we develop *Proteina*, a new large-scale flow-based protein backbone generator that utilizes hierarchical fold class labels for conditioning and relies on a tailored scalable transformer architecture with up to $5\times$ as many parameters as previous models. To meaningfully quantify performance, we introduce a new set of metrics that directly measure the distributional similarity of generated proteins with reference sets, complementing existing metrics. We further explore scaling training data to millions of synthetic protein structures and explore improved training and sampling recipes adapted to protein backbone generation. This includes fine-tuning strategies like LoRA for protein backbones, new guidance methods like classifier-free guidance and autoguidance for protein backbones, and new adjusted training objectives. Proteina achieves state-of-the-art performance on de novo protein backbone design and produces diverse and designable proteins at unprecedented length, up to 800 residues. The hierarchical conditioning offers novel control, enabling high-level secondary-structure guidance as well as low-level fold-specific generation.

Molecular property prediction is one of the fastest-growing applications of deep learning with critical real-world impacts. Although the 3D molecular graph structure is necessary for models to achieve strong performance on many tasks, it is infeasible to obtain 3D structures at the scale required by many real-world applications. To tackle this issue, we propose to use existing 3D molecular datasets to pre-train a model to reason about the geometry of molecules given only their 2D molecular graphs. Our method, called 3D Infomax, maximizes the mutual information between learned 3D summary vectors and the representations of a graph neural network (GNN). During fine-tuning on molecules with unknown geometry, the GNN is still able to produce implicit 3D information and uses it for downstream tasks. We show that 3D Infomax provides significant improvements for a wide range of properties, including a 22% average MAE reduction on QM9 quantum mechanical properties. Moreover, the learned representations can be effectively transferred between datasets in different molecular spaces.

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

MOTIVATION: Classifying proteins into functional families can improve our understanding of protein function and can allow transferring annotations within one family. For this, functional families need to be 'pure', i.e., contain only proteins with identical function. Functional Families (FunFams) cluster proteins within CATH superfamilies into such groups of proteins sharing function. 11% of all FunFams (22 830 of 203 639) contain EC annotations and of those, 7% (1526 of 22 830) have inconsistent functional annotations. RESULTS: We propose an approach to further cluster FunFams into functionally more consistent sub-families by encoding their sequences through embeddings. These embeddings originate from language models transferring knowledge gained from predicting missing amino acids in a sequence (ProtBERT) and have been further optimized to distinguish between proteins belonging to the same or a different CATH superfamily (PB-Tucker). Using distances between embeddings and DBSCAN to cluster FunFams and identify outliers, doubled the number of pure clusters per FunFam compared to random clustering. Our approach was not limited to FunFams but also succeeded on families created using sequence similarity alone. Complementing EC annotations, we observed similar results for binding annotations. Thus, we expect an increased purity also for other aspects of function. Our results can help generating FunFams; the resulting clusters with improved functional consistency allow more reliable inference of annotations. We expect this approach to succeed equally for any other grouping of proteins by their phenotypes. AVAILABILITY AND IMPLEMENTATION: Code and embeddings are available via GitHub: https://github.com/Rostlab/FunFamsClustering. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: Coevolutionary sequence analysis has become a commonly used technique for de novo prediction of the structure and function of proteins, RNA, and protein complexes. We present the EVcouplings framework, a fully integrated open-source application and Python package for coevolutionary analysis. The framework enables generation of sequence alignments, calculation and evaluation of evolutionary couplings (ECs), and de novo prediction of structure and mutation effects. The combination of an easy to use, flexible command line interface and an underlying modular Python package makes the full power of coevolutionary analyses available to entry-level and advanced users. AVAILABILITY AND IMPLEMENTATION: https://github.com/debbiemarkslab/evcouplings.

BACKGROUND: Predicting protein function and structure from sequence is one important challenge for computational biology. For 26 years, most state-of-the-art approaches combined machine learning and evolutionary information. However, for some applications retrieving related proteins is becoming too time-consuming. Additionally, evolutionary information is less powerful for small families, e.g. for proteins from the Dark Proteome. Both these problems are addressed by the new methodology introduced here. RESULTS: We introduced a novel way to represent protein sequences as continuous vectors (embeddings) by using the language model ELMo taken from natural language processing. By modeling protein sequences, ELMo effectively captured the biophysical properties of the language of life from unlabeled big data (UniRef50). We refer to these new embeddings as SeqVec (Sequence-to-Vector) and demonstrate their effectiveness by training simple neural networks for two different tasks. At the per-residue level, secondary structure (Q3 = 79% ± 1, Q8 = 68% ± 1) and regions with intrinsic disorder (MCC = 0.59 ± 0.03) were predicted significantly better than through one-hot encoding or through Word2vec-like approaches. At the per-protein level, subcellular localization was predicted in ten classes (Q10 = 68% ± 1) and membrane-bound were distinguished from water-soluble proteins (Q2 = 87% ± 1). Although SeqVec embeddings generated the best predictions from single sequences, no solution improved over the best existing method using evolutionary information. Nevertheless, our approach improved over some popular methods using evolutionary information and for some proteins even did beat the best. Thus, they prove to condense the underlying principles of protein sequences. Overall, the important novelty is speed: where the lightning-fast HHblits needed on average about two minutes to generate the evolutionary information for a target protein, SeqVec created embeddings on average in 0.03 s. As this speed-up is independent of the size of growing sequence databases, SeqVec provides a highly scalable approach for the analysis of big data in proteomics, i.e. microbiome or metaproteome analysis. CONCLUSION: Transfer-learning succeeded to extract information from unlabeled sequence databases relevant for various protein prediction tasks. SeqVec modeled the language of life, namely the principles underlying protein sequences better than any features suggested by textbooks and prediction methods. The exception is evolutionary information, however, that information is not available on the level of a single sequence.