Platforma

Platforma is the end-to-end bioinformatics software application built by MiLaboratories for high-throughput antibody, TCR, and peptide discovery campaigns.  It covers both in vitro discovery (phage and yeast display) and in vivo discovery (immunization and single-cell), running clonotyping, enrichment analysis, liability flagging, and structural clustering in one governed, white-box environment.

Platforma allows wet-lab scientists to execute complex computational workflows without a dedicated computational biology team and trasnform raw NGS reads to a ranked panel of diverse, developable leads.

Built on MiXCR, the most widely used tool for immune repertoire profiling, Platforma provides purpose-built, end-to-end discovery solutions for

• In vitro antibody discovery — deconvolute phage and yeast display libraries, track enrichment across panning rounds, and separate true binders from fast-growing passenger clones.
• In vivo antibody discovery — resolve native heavy/light chain pairings, trace somatic hypermutation to germline, and integrate bulk, single-cell, and multiomics data into one decision matrix.

How to use Platforma

Platforma is deployed on your own infrastructure (HPC or private cloud). Users can choose from different biological blocks or build custom blocks:

• Annotation and clonotyping — Process 10x Genomics VDJ outputs, single B-cell sequencing data, and VHH/scFv/Fab display libraries with the built-in MiXCR engine to extract distinct clonotypes and assign V(D)J genes instantly
• Clonotype clustering at scale — Handle millions of B-cell repertoire sequences without a dedicated computational biology team. Group sequences into distinct functional families using the MMSeqs2 engine—by CDRs, framework regions, or full-length sequences—to guarantee a structurally diverse lead panel.
• Library characterization — assess starting-library diversity and structural integrity with sequence-level metrics across selection rounds
• Lineage tracing and SHM trees — build somatic hypermutation trees to track affinity maturation from common germline ancestors, with the phylogenetic relationships fully visible rather than hidden in black-box logic
• Enrichment analysis — calculate fold-change across rounds, using negative controls to distinguish antigen-specific binders from fast-growing passenger clones.
• Developability & liability flagging — Automatically rank antibody candidates by developability, flagging structural and sequence-based risks directly on the sequence before costly wet-lab validation.
• Structure modeling — generate 3D antibody models to assess paratope surfaces and support structure-informed selection.
• Lead prioritization — triangulate enrichment scores, liability flags, and clustering data in one workspace, visualized on an interactive antibody-space map

Core Features of Platforma:

• V(D)J assembly and clonotyping with MiXCR error correction
• Enrichment / fold-change analysis across panning rounds, with negative-control tracking
• Somatic hypermutation (SHM) lineage trees to germline ancestors
• Antigen specificity mapping — overlay surface markers, flow cytometry, and antigen-binding scores onto repertoire data to validate function
• Interactive data visualization: A suite of dynamic charts and interactive UMAP to visualize the entire antibody space
• Paratope and 3D structural clustering by predicted epitope
• Sequence and structural liability annotation for developability de-risking
• Lead panel selection integrating enrichment, liabilities, clustering, and functional data

Key use cases

• In vitro antibody discovery: Mine selection outputs from phage display and yeast display technologies to pinpoint high-affinity, stably enriched binders
• In vivo antibody discovery: Discover rare, high-affinity therapeutic leads by mining natural immune repertoires, track somatic hypermutation (SHM) with robust phylogenetic lineage trees, and integrate single-cell multi-omics with native heavy/light chain pairing
• Peptide discovery: Track enrichment across peptide display libraries to identify diverse, developable leads
• Translational research: Decode the adaptive immune repertoire, map TCR/BCR repertoire signals directly to patient outcomes, and compare diverse clinical cohorts
• Multi-omics analysis: Integrate transcriptomics and immune repertoires to uncover unique cell-type markers and enriched functional pathways

Advantages

• End-to-end, no-code workflow: A comprehensive solution that seamlessly drives the entire discovery pipeline in an interactive user interface.
• Billion-scale data processing: Engineered for massive throughput, effortlessly analyzing billions of sequences without introducing computational bottlenecks.
• Intelligent lead selection: Lead selection filters, scores, and diversifies. Using a composite rank, it avoids naive "Top-N" sorting that yields redundant clones. Instead, it selects the best representative from each clonal family —delivering leads that are best-in-class, structurally diverse, and highly developable in one unified step
• It adapts to your biology, not the other way around. Non-standard formats (VHH, bispecifics like knob-in-hole, tri-specifics, custom scaffolds) and unusual experimental designs (DMS, pH-switch selection, multi-arm panning, Tite-Seq) are configured to any protocol. If the block doesn't exist yet, our scientists build it.
• Transparent — no black boxes. Every filter, weight, and ranking rule is shown and editable; block logic is open-source and auditable, on published gold-standard engines.
• Deploy where the data lives. Your cloud, on-prem. Enterprise-grade data sovereignty is a foundational feature