Dirichlet Flow Matching with Applications to DNA Sequence Design.

Discrete diffusion or flow models could enable faster and more controllable sequence generation than autoregressive models. We show that na\"ive linear flow matching on the simplex is insufficient toward this goal since it suffers from discontinuities in the training target and further pathologies. To overcome this, we develop Dirichlet flow matching on the simplex based on mixtures of Dirichlet distributions as probability paths. In this framework, we derive a connection between the mixtures' scores and the flow's vector field that allows for classifier and classifier-free guidance. Further, we provide distilled Dirichlet flow matching, which enables one-step sequence generation with minimal performance hits, resulting in $O(L)$ speedups compared to autoregressive models. On complex DNA sequence generation tasks, we demonstrate superior performance compared to all baselines in distributional metrics and in achieving desired design targets for generated sequences. Finally, we show that our classifier-free guidance approach improves unconditional generation and is effective for generating DNA that satisfies design targets. Code is available at https://github.com/HannesStark/dirichlet-flow-matching.

Benchmarking AlphaFold-enabled molecular docking predictions for antibiotic discovery.

Efficient identification of drug mechanisms of action remains a challenge. Computational docking approaches have been widely used to predict drug binding targets; yet, such approaches depend on existing protein structures, and accurate structural predictions have only recently become available from AlphaFold2. Here, we combine AlphaFold2 with molecular docking simulations to predict protein-ligand interactions between 296 proteins spanning Escherichia coli's essential proteome, and 218 active antibacterial compounds and 100 inactive compounds, respectively, pointing to widespread compound and protein promiscuity. We benchmark model performance by measuring enzymatic activity for 12 essential proteins treated with each antibacterial compound. We confirm extensive promiscuity, but find that the average area under the receiver operating characteristic curve (auROC) is 0.48, indicating weak model performance. We demonstrate that rescoring of docking poses using machine learning-based approaches improves model performance, resulting in average auROCs as large as 0.63, and that ensembles of rescoring functions improve prediction accuracy and the ratio of true-positive rate to false-positive rate. This work indicates that advances in modeling protein-ligand interactions, particularly using machine learning-based approaches, are needed to better harness AlphaFold2 for drug discovery.

Light attention predicts protein location from the language of life.

Summary: Although knowing where a protein functions in a cell is important to characterize biological processes, this information remains unavailable for most known proteins. Machine learning narrows the gap through predictions from expert-designed input features leveraging information from multiple sequence alignments (MSAs) that is resource expensive to generate. Here, we showcased using embeddings from protein language models for competitive localization prediction without MSAs. Our lightweight deep neural network architecture used a softmax weighted aggregation mechanism with linear complexity in sequence length referred to as light attention. The method significantly outperformed the state-of-the-art (SOTA) for 10 localization classes by about 8 percentage points (Q10). So far, this might be the highest improvement of just embeddings over MSAs. Our new test set highlighted the limits of standard static datasets: while inviting new models, they might not suffice to claim improvements over the SOTA. Availability and implementation: The novel models are available as a web-service at http://embed.protein.properties. Code needed to reproduce results is provided at https://github.com/HannesStark/protein-localization. Predictions for the human proteome are available at https://zenodo.org/record/5047020. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

Diagnostic performance of [18F] FDG PET-CT compared to bone scintigraphy for the detection of bone metastases in lung cancer patients.

BACKGROUND: Accurate staging of lung cancer is essential for effective patient management and selection of appropriate therapeutic strategy. The aim of this paper was to compare the value of bone scintigraphy and FDG PET-CT for detecting bone metastases in lung cancer patients and the impact of these modalities in disease staging. METHODS: One hundred sixty-four lung cancer patients who had undergone both FDG PET-CT and bone scintigraphy within 14 days were included into this study. The analysis of FDG PET-CT and bone scintigraphy was carried out patient- and lesion-based. RESULTS: One hundred twenty-one patients were negative and 43 patients positive for bone metastases. FDG PET-CT found bone metastases in 42/43 patients and bone scintigraphy in 38/43 patients. Sensitivity, specificity and accuracy of FDG PET-CT and bone scintigraphy for detecting bone metastases were 97.7%, 100% and 99.4%, and 87.8%, 97.5% and 94.2%, respectively. FDG PET-CT identified 430 bone metastases and bone scintigraphy 246 bone metastases. Skull was the only region where bone scintigraphy identified more lesions than FDG PET-CT. Based on both scintigraphic modalities disagreement concerning disease stage was found in 3 patients. CONCLUSION: FDG PET-CT yielded a higher sensitivity, specificity and accuracy than bone scintigraphy for identifying bone metastases in lung cancer patients. FDG PET-CT thus can be recommended for initial staging of lung cancer patients without applying bone scintigraphy for the detection of bone metastases.