2013 was a big year: Andy Murray clinched the Wimbledon title, NASA’s Curiosity Rover discovered water-bearing minerals on Mars, and ‘twerk’ and ‘selfie’ made their way into the dictionary, something equally significant happened—the birth of BLOPIG.com. Intrigued by how the group has changed over the last decade, I started on a journey to unearth the some of the publications from then till now. Questioning their focus, methods, and evolution of the groups’ research over the past decade. This blog post is what I found.
While delving into each publication of the past decade genuinely seemed like an interesting idea, the imminent threat to my PhD progress forced me to adopt the most 2023-appropriate approach: outsourcing the task to AI. After collecting abstracts from all the group’s papers, I enlisted the help of everyone’s’ favourite hallucinator to summarise the works and (hopefully) highlight the shifts in their research.
So after a relatively long, sequence of prompts, this is (apparently) what we do?
In the early years, the research group focused on studying the evolution of proteins and understanding their long-term dynamics, including changes in structure and function. They explored the relationship between protein ages and various properties, such as sequence, structure, and function. They observed differences between newly evolved and ancient protein structures, suggesting a less elaborate structure for newly evolved superfamilies.
In the later years, the research group expanded their focus to include the investigation of local interaction network patterns in protein-protein interfaces. They identified preferred chemical patterns of closely packed residues at the interface and introduced the iScore method to assess interface predictions. Additionally, they examined the structure and prediction of protein loops, finding that the distribution of loop span is independent of the number of residues in the loop. They proposed a random fluctuation model based on the Maxwell-Boltzmann distribution to describe the distribution of loop span and investigated its effect on protein loop structure prediction. The group also focused on antibodies, developing methods to annotate antibody residues likely to be in contact with antigens. They used antibody-specific statistics to enhance their annotation method and achieve high-quality results in a short time .
Overall, the research group expanded their interests from studying protein evolution to include protein-protein interfaces, protein loops, and antibodies, while still maintaining a focus on understanding the relationship between structure, function, and evolution.
My initial attempt left me dissatisfied. Was this a task to big? So, with a refined process and some actual reading on my part. Let’s dive into the scientific odyssey of OPIG.
OPIG began its second decade in 2013 with a primary focus on protein evolution. This involved a exploration that went beyond the realms of sequence-level mutations, extending into the structural and functional dimensions of proteins. The annotation of superfamily age provided a robust foundation, allowing the group to discern disparities between newly evolved and ancient protein structures. This foundational work established a toolkit for analysing protein populations comprehensively. The initial years saw a concentration on protein-protein interactions, introducing innovative methods such as the iScore method for high-specificity interface predictions. Simultaneously, challenges in protein loop structure prediction were addressed, refining difficulty metrics based on loop span. The development of tools like Antibody i-Patch showcased a practical application, providing insights for artificial affinity maturation in antibody engineering.
As the years progressed, the group’s interests expanded horizontally and vertically. From delving into vascular tissue engineering and the structural analysis of centriole-forming proteins to tool development for systematic protein-ligand structure analysis, the research landscape became increasingly diverse. Notable contributions included WONKA for enhanced protein-ligand structure analysis and studies on kinases, peptide-MHC binding affinity prediction, and identifying structural bridges between protein folds.
The groups research expanded into investigations into molecular interactions, antibody structures, and computational modelling. Novel tools like ABodyBuilder and SAbPred streamlined antibody modelling and predicted antibody properties, contributing to therapeutic antibody design. The horizon broadened to encompass studies on HIV-1 replication dynamics, systematic biomolecular functional motions, and a network similarity score (Netdis) for enriched understanding of complex biological processes.
The latter half of the decade witnessed an expansion into the realms of next-gen sequencing, structural information integration, and crystallographic fragment screening. Innovations like the PanDDA method for studying protein dynamics and Sphinx for loop modeling in antibodies reflected a continuous commitment to pushing the boundaries of structural biology.
The group’s interests in 2018 ranged from next-gen sequencing challenges and molecular simulations’ reliability to bioinformatics tools like pyHVis3D and ANARCI. Epigenetic modifications, loop prediction algorithms, and innovative approaches in vascular tissue engineering demonstrated a multifaceted exploration of complex biological phenomena. The next year saw Ligity for virtual screening, Bayesian optimisation, and molecular simulations exploring T cell receptor dynamics. Studies on MHC class II stabilisation mechanisms, homology modelling of kinases, and algorithms like RFQAmodel and SCALOP highlighted the group’s dedication to model quality assessment and guidelines for antibody canonical forms. The groups interested continued to develop. From areas like antibody therapeutics tracking, machine learning for protein-ligand binding, conformer sampling, and B-cell receptor profiling in COVID-19 patients in 2020 to quantum computing’s potential in computational biology, conformational entropy prediction, and advancements in protein folding using quantum annealing in 2021.
In 2022, the research spanned hydrogen deuterium exchange mass spectrometry, Hepatitis B virus mutations, machine learning scoring functions, TRIM33 isoforms, SARS-CoV-2 RNA nuclease complex, computational developability assessment, antibody repertoires, and language models for antibody sequences. Updates to SAbDab, antibody-specific predictive models, and developments like AbLang and multiplex brain state graph models reflected ongoing advancements. The momentum continued into 2023, encapsulating critical evaluations of structure-based virtual screening models, improved antibody epitope prediction with SPACE2, challenges to conventional fragment-based drug design, anti-chemokine peptides from tick evasins, Paragraph for structure-based paratope prediction, ImmuneBuilder for accurate immune receptor structure prediction, CoPriNet for predicting compound prices, Fragment Network for merging in crystallographic screening, and PointVS addressing biases in machine learning-based scoring functions. Additionally, a curriculum-learning-inspired procedure in deep reinforcement learning showed promise in enhancing diversity in de novo molecular design.
Methodologically, the group’s journey began with pioneering computational tools addressing specific challenges. ABangle introduced structural intricacies, while DenseNet showcased the power of machine learning in structure-based virtual screening. The integration of quantum computing, Bayesian optimisation, and deep learning reflected a commitment to leveraging state-of-the-art methodologies. Emerging tools like SPACE2 and ImmuneBuilder demonstrated a synergy between traditional and contemporary approaches, highlighting the group’s agility in tackling the ever-expanding complexity of molecular sciences.
OPIG’s two-decade-long evolution traces a trajectory from protein evolution to a diverse portfolio covering molecular interactions, structural biology, computational modelling, and the integration of cutting-edge technologies. This journey is marked by a commitment to innovation, a multidisciplinary approach, and an agility to embrace emerging methodologies, defining OPIG’s distinctive place in the ever-evolving field of molecular sciences.
There you have it. 10 years in 8 paragraphs.
