Biology subjects

Kocikowski, M.

Publications and source records attributed to Kocikowski, M..

4 recordsLinked to original sources

bioRxiv2026

SpeciefAI: Multi-species mRNA-level Antibody Framework Generation using Transformers

MotivationEncoding antibodies (Abs) and nanobodies (Nbs) as mRNA enables in vivo production of therapeutic proteins. However, this approach requires meeting two species-dependent requirements: the mRNA encoding must support efficient expression in the host species, and the encoded protein sequence must resemble the natural Ab repertoire of the recipient species to minimize immunogenicity. These requirements motivate species-conditioned generative models for joint mRNA and protein design. ResultsWe propose SpeciefAI a transformer-based model for multi-species Ab and Nb species sequence-harmonisation by generation of novel Framework Regions (FRs) tailored to input Complementarity-Determining Regions (CDRs). Our model works directly in the mRNA space and learns the correspondence between FRs and CDRs in six species. The model is capable of generating sequences with a highly similar distribution to natural sequences and a mean absolute difference in codon adaptation index (CAI) of 0.013 and 0.033 for humans and dogs respectively. We show that the generated human sequences are highly human (0.95 T20 score) and canine sequences highly canine (0.95 cT20 score). We furthermore demonstrate that we can generate diverse candidate sequences using our method. Availability and ImplementationSource code is available on https://github.com/Dominko/SpeciefAI. OAS and COGNANO data are publicly available on https://opig.stats.ox.ac.uk/webapps/oas/ and https://cognanous.com/datasets/vhh-corpus (preprocessed versions available upon request). Canine data is available on https://zenodo.org/records/18301526.

bioinformatics↗

bioRxiv2025

DoggifAI: a transformer based approach for antibodycaninisation

Antibody translation across species offers a compelling strategy to extend the vast and expensive investments in human therapeutic antibodies to veterinary oncology, with applications in both veterinary medicine and comparative oncology. While precise, low-immunogenic treatments are essential for canine cancer care, traditional species conversion methods rely on ad hoc bioinformatics modifications. These methods often implicitly decouple the framework (FR) and complementarity-determining regions (CDRs), ignoring how structural changes in FRs can affect the conformation and function of CDRs. This can compromise binding specificity and require costly high-throughput in vitro screening. To address this, we present DoggifAI, a transformer model that translates non-canine antibody sequences into canine ones by generating species-appropriate framework regions (FRs) based on desired CDRs. This allows the model to better preserve structural compatibility between FRs and CDRs. The model is pretrained in a T5-style text-to-text denoising task on a large multispecies antibody dataset, which allows further finetuning on a much smaller species-specific dataset. DoggifAI generates highly canine-like antibodies and shows promising results in preserving binding specificity. To support further progress in this field, we also release a curated dataset of over 430,000 unique canine antibody chain sequences, significantly expanding the public sequence repertoire. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=127 SRC="FIGDIR/small/656573v1_ufig1.gif" ALT="Figure 1"> View larger version (27K): org.highwire.dtl.DTLVardef@13c10adorg.highwire.dtl.DTLVardef@6afca6org.highwire.dtl.DTLVardef@1f11961org.highwire.dtl.DTLVardef@1b9563a_HPS_FORMAT_FIGEXP M_FIG Graphical Abstract C_FIG HighlightsO_LIWe show that transformer models are capable of generating plausible antibody framework regions based on CDRs C_LIO_LIWe show that resulting framework regions are highly recognisable as coming from the desired species C_LIO_LIWe show promising results for the retention of binding specificity when translating antibody sequences in this way C_LIO_LIWe release a large, high-quality dataset of canine antibody sequences to support future research C_LI

bioinformatics↗

bioRxiv2025

Expanding the definition of MHC Class I peptide binding promiscuity to support vaccine discovery across cancers with CARMEN

Promiscuity in T-cell antigen landscapes refers to the dual flexibility of peptides binding multiple MHC alleles and MHC alleles presenting diverse arrays of peptides. By understanding how neoantigens are shared across varied HLA backgrounds, promiscuity analysis can inform the selection of cancer-vaccine targets that reach a wider segment of the population and help refine patient stratification for diverse immunotherapies. We expand the concept of promiscuity to encompass peptides, MHC alleles, individuals, populations, and genomic regions. Our CARMEN database release harmonizes data from 72 publications (2,323 samples) across tissue types, with a focus on cancer. Using Gibbs clustering and dimensionality reduction (UMAP), we systematically map promiscuity and immunological versatility across these biological levels. Gene and mutation analysis reveals recurrent cancer mutations in highly promiscuous genomic regions, highly mutated cancer genes that avoid presented regions, and sheds light on genomic regions important to response to immunotherapy.

bioinformatics↗

bioRxiv2024

Barking Up the Right Tree: Immune Checkpoint Signatures of Human and Dog Cancers

In the quest for improved therapeutics targeting immune checkpoints (ICs), we turn to spontaneously developing dog (canine) cancers, which are unique models that genetically and clinically mirror human equivalents. Despite its potential, canine cancer immunology remains largely unexplored. Here, we examine the RNA-seq-based expression of 44 ICs across 14 canine cancer types and an extensive human dataset. We unveil diverse canine IC expression patterns and unique human IC signatures that reflect the histological type and primary site of cancer. We uncover a striking similarity between canine brain cancers, osteosarcoma, and their human counterparts, identifying them as prospective immunotherapy models. Four ICs--CD160, A2AR, NKG2A, and OX40--are key to the differences observed between species. Moreover, individual patient IC signatures exhibit varying alignment with their respective cancer types, a finding with profound implications for personalized human therapy. This exploration illuminates new aspects of canine and human cancer immunology, setting the stage for discoveries at their crossroads.

cancer biology↗