EMERALD-UI: An interactive web application to unveil novel protein biology hidden in the suboptimal-alignment space

Life over the past four billion years has been shaped by proteins and their capacity to assemble into three dimensional conformations. Protein sequence alignments have been the enabling technology for exploring the evolution and functional adaptation of proteins across the tree of life. Recent advancements in scaling the prediction of three dimensional protein structures from primary sequence alone, revealed that different modes of conservation and function operate on the sequence and structure level. This difference in protein conservation patterns and their underlying functional change that could emerge in suboptimal alignment configurations is often ignored in optimal protein alignment approaches. We introduce EMERALD-UI, an open-source interactive web application which is designed to reveal unexplored biology by visualising stable structural conformations or protein regions hidden in the suboptimal alignment space. Availability: EMERALD-UI is available at https://algbio.github.io/emerald-ui/. Contact: hdrost001@dundee.ac.uk or alexandru.tomescu@helsinki.fi.

💡 Research Summary

The paper introduces EMERALD‑UI, a browser‑based interactive web application designed to expose biologically relevant information hidden in the suboptimal alignment space of protein sequences. Traditional alignment methods focus on a single optimal alignment, assuming that the lowest‑cost path corresponds to the most biologically meaningful pairing. While this assumption holds for highly similar proteins, it becomes problematic for divergent sequences where thousands of alternative alignments exist with only slightly higher scores. The authors previously released EMERALD, a command‑line tool that enumerates suboptimal alignments using a directed‑acyclic graph (DAG) representation of the Needleman‑Wunsch dynamic programming matrix, and extracts “safety windows” – contiguous intervals that are conserved across a user‑defined proportion (α) of all alignments within a cost tolerance (Δ).

EMERALD‑UI builds on this algorithmic foundation by porting the core EMERALD engine to WebAssembly, allowing it to run entirely client‑side without any server installation. The UI accepts protein sequences via FASTA upload, manual entry, PDB/CIF files, or direct UniProt search. Users can adjust Δ (suboptimal depth) and α (safety threshold) with sliders, and also fine‑tune the scoring matrix (BLOSUM45‑90, PAM30‑250, IDENTITY) and gap penalties. The application then visualizes the full alignment DAG as an interactive D3.js graph: the optimal path appears in blue, while alternative suboptimal paths are shown in red/orange. Hovering over any segment reveals the underlying residues, their indices, and the fraction of alignments that traverse that segment. Safety windows are highlighted in green, and can be selected for further inspection.

Structural mapping is performed using Mol*; safety‑window residues are coloured on the 3‑D protein model, which is fetched in real time from the AlphaFold or UniProt APIs. EMERALD‑UI also provides direct hyperlinks to downstream resources such as KEGG, FoldSeek, and LigandDB, enabling users to explore pathway context, structural homologues, or ligand‑binding potential of the identified regions. All outputs—raw alignment text, selected windows, and the visual graph—can be exported as PNG/JPEG/SVG, plain text, or a shareable URL that reproduces the exact state of the analysis.

The authors benchmarked the system on the SwissProt database, demonstrating that for highly divergent sequence pairs, safety windows often cover a small fraction of the total alignment yet retain a high proportion of structurally stable residues. This confirms that suboptimal alignments can reveal conserved structural cores missed by a single optimal alignment. Moreover, the tool is positioned to aid protein‑complex analysis: by selecting alternative alignment paths, researchers can hypothesize new interface configurations, potentially uncovering novel interaction sites or druggable pockets.

Technical strengths include a lightweight, serverless architecture, real‑time responsiveness even for large inputs, and extensive configurability of evolutionary models. The integration of interactive graph exploration, 3‑D visualization, and seamless connection to major bioinformatics databases makes EMERALD‑UI a comprehensive platform for both exploratory research and hypothesis‑driven investigations.

In summary, EMERALD‑UI extends the concept of alignment safety from a command‑line utility to an accessible, feature‑rich web application. By allowing users to interrogate the full suboptimal alignment landscape, map conserved regions onto up‑to‑date structural models, and immediately link to functional annotations, the tool opens new avenues for discovering hidden functional motifs, refining structure‑function relationships, and guiding downstream experimental design.