The rapid advancement of LLMs has led to widespread adoption across various domains, but it has also raised concerns about data security and privacy, particularly with publicly available and commercially operated platforms. Given their high computational demands, cloud environments are the obvious choice for deployment. As a result, organizations are increasingly deploying LLMs in confined cloud environments to protect sensitive data while leverazing scalable cloud resources. However, deploying LLMs in cloud environments remains a complex and time-consuming process that requires specialized skills and expertise in various areas, such as infrastructure management, resource allocation, and model setup. Testing and comparing LLMs to select the appropriate one is particularly challenging as different models are trained for different purposes, making the direct comparison nontrivial. Furthermore, differences in model architectures, training data, and fine-tuning strategies make objective evaluation difficult, limiting the effectiveness of traditional benchmarking approaches. To address these challenges, we present a cloud-native system that automates both the deployment and evaluation of LLMs. Our contributions are twofold: (i) we automate the provisioning and deployment of LLMs on various cloud platforms to stream-line infrastructure setup, and (ii) we develop a lightweight evaluation framework that leverages the LLM-as-a-Judge approach, where an independent LLM systematically assesses and compares different models based on predefined evaluation criteria. Our ongoing work aims to optimize LLM deployment by selecting cost-efficient cloud resources. We are also enhancing the evaluation framework with diverse prompts, broader metrics, and cross-model validation for fair, reproducible benchmarking.

MOTIVATION: Machine learning-based scoring functions (MLBSFs) have been found to exhibit inconsistent performance on different benchmarks and be prone to learning dataset bias. For the field to develop MLBSFs that learn a generalizable understanding of physics, a more rigorous understanding of how they perform is required. RESULTS: In this work, we compared the performance of a diverse set of popular MLBSFs (RFScore, SIGN, OnionNet-2, Pafnucy, and PointVS) to our proposed baseline models that can only learn dataset biases on a range of benchmarks. We found that these baseline models were competitive in accuracy to these MLBSFs in almost all proposed benchmarks, indicating these models only learn dataset biases. Our tests and provided platform, ToolBoxSF, will enable researchers to robustly interrogate MLBSF performance and determine the effect of dataset biases on their predictions. AVAILABILITY AND IMPLEMENTATION: https://github.com/guydurant/toolboxsf.

MOTIVATION: The sharing of macromolecular structural information online by scientists is predominantly performed via 2D static images, since the embedding of interactive 3D structures in webpages is non-trivial. Whilst the technologies to do so exist, they are often only implementable with significant web coding experience. RESULTS: Michelaɴɢʟo is an accessible and open-source web-based application that supports the generation, customization and sharing of interactive 3D macromolecular visualizations for digital media without requiring programming skills. A PyMOL file, PDB file, PDB identifier code or protein/gene name can be provided to form the basis of visualizations using the NGL JavaScript library. Hyperlinks that control the view can be added to text within the page. Protein-coding variants can be highlighted to support interpretation of their potential functional consequences. The resulting visualizations and text can be customized and shared, as well as embedded within existing websites by following instructions and using a self-contained download. Michelaɴɢʟo allows researchers to move away from static images and instead engage, describe and explain their protein to a wider audience in a more interactive fashion. AVAILABILITY AND IMPLEMENTATION: Michelaɴɢʟo is hosted at michelanglo.sgc.ox.ac.uk. The Python code is freely available at https://github.com/thesgc/MichelaNGLo, along with documentations about its implementation.

MOTIVATION: Protein ubiquitination plays a central role in important cellular machineries such as protein degradation or chromatin-mediated signaling. With the recent discovery of the first potent ubiquitin-specific protease inhibitors, and the maturation of proteolysis targeting chimeras as promising chemical tools to exploit the ubiquitin-proteasome system, protein target classes associated with ubiquitination pathways are becoming the focus of intense drug-discovery efforts. RESULTS: We have developed UbiHub, an online resource that can be used to visualize a diverse array of biological, structural and chemical data on phylogenetic trees of human protein families involved in ubiquitination signaling, including E3 ligases and deubiquitinases. This interface can inform target prioritization and drug design, and serves as a navigation tool for medicinal chemists, structural and cell biologists exploring ubiquitination pathways. AVAILABILITY AND IMPLEMENTATION: https://ubihub.thesgc.org.

UNLABELLED: The rapidly increasing research activity focused on chromatin-mediated regulation of epigenetic mechanisms is generating waves of data on writers, readers and erasers of the histone code, such as protein methyltransferases, bromodomains or histone deacetylases. To make these data easily accessible to communities of research scientists coming from diverse horizons, we have created ChromoHub, an online resource where users can map on phylogenetic trees disease associations, protein structures, chemical inhibitors, histone substrates, chromosomal aberrations and other types of data extracted from public repositories and the published literature. The interface can be used to define the structural or chemical coverage of a protein family, highlight domain architectures, interrogate disease relevance or zoom in on specific genes for more detailed information. This open-access resource should serve as a hub for cell biologists, medicinal chemists, structural biologists and other navigators that explore the biology of chromatin signalling. AVAILABILITY: http://www.thesgc.org/chromohub/.

SUMMARY: The structural genomics of histone tail recognition web server is an open access resource that presents within mini articles all publicly available experimental structures of histone tails in complex with human proteins. Each article is composed of interactive 3D slides that dissect the structural mechanism underlying the recognition of specific sequences and histone marks. A concise text html-linked to interactive graphics guides the reader through the main features of the interaction. This resource can be used to analyze and compare binding modes across multiple histone recognition modules, to evaluate the chemical tractability of binding sites involved in epigenetic signaling and design small molecule inhibitors. AVAILABILITY: http://www.thesgc.org/resources/histone_tails/ CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.