TIS Transformer: remapping the human proteome using deep learning

Characterization of shared neoantigens landscape in Mismatch Repair Deficient Endometrial Cancer

Endometrial cancer (EC) with Mismatch Repair deficiency (MMRd) is characterized by the accumulation of insertions/deletions at microsatellite sites. These mutations lead to the synthesis of frameshift peptides (FSPs) that represent tumor-specific neoantigens (nAg) proved to be shared across patients/tumors with MMRd. In this study, we explored the feasibility of a nAg-based cancer vaccination design in EC with MMRd. We adopted a whole exome sequencing approach and ad hoc bioinformatics pipelines to characterize FSPs in 35 patients with EC. A mean of 146 mutated mononucleotide repeats (MNRs) was identified with enrichment in the patients' group with MLH1 impairment. A high coverage emerged from the comparative analysis of the EC FSPs with the content of the previously validated NOUS-209 vaccine. We obtained pieces of evidence of FSPs translation as expressed proteins from Ribo-seq, supporting the potential as the target of vaccination. The development of a nAgs-based vaccine strategy in MMRd EC may be further explored.

Deep learning to decode sites of RNA translation in normal and cancerous tissues

The biological process of RNA translation is fundamental to cellular life and has wide-ranging implications for human disease. Yet, accurately delineating the variation in RNA translation represents a significant challenge. Here, we develop RiboTIE, a transformer model-based approach to map global RNA translation. We find that RiboTIE offers unparalleled precision and sensitivity for ribosome profiling data. Application of RiboTIE to normal brain and medulloblastoma cancer samples enables high-resolution insights into disease regulation of RNA translation.

Untranslated regions (UTRs) are a potential novel source of neoantigens for personalised immunotherapy

Background

Neoantigens, mutated tumour-specific antigens, are key targets of anti-tumour immunity during checkpoint inhibitor (CPI) treatment. Their identification is fundamental to designing neoantigen-directed therapy. Non-canonical neoantigens arising from the untranslated regions (UTR) of the genome are an overlooked source of immunogenic neoantigens. Here, we describe the landscape of UTR-derived neoantigens and release a computational tool, PrimeCUTR, to predict UTR neoantigens generated by start-gain and stop-loss mutations.

Methods

We applied PrimeCUTR to a whole genome sequencing dataset of pre-treatment tumour samples from CPI-treated patients (n = 341). Cancer immunopeptidomic datasets were interrogated to identify MHC class I presentation of UTR neoantigens.

Results

Start-gain neoantigens were predicted in 72.7% of patients, while stop-loss mutations were found in 19.3% of patients. While UTR neoantigens only accounted 2.6% of total predicted neoantigen burden, they contributed 12.4% of neoantigens with high dissimilarity to self-proteome. More start-gain neoantigens were found in CPI responders, but this relationship was not significant when correcting for tumour mutational burden. While most UTR neoantigens are private, we identified two recurrent start-gain mutations in melanoma. Using immunopeptidomic datasets, we identify two distinct MHC class I-presented UTR neoantigens: one from a recurrent start-gain mutation in melanoma, and one private to Jurkat cells.

Conclusion

PrimeCUTR is a novel tool which complements existing neoantigen discovery approaches and has potential to increase the detection yield of neoantigens in personalised therapeutics, particularly for neoantigens with high dissimilarity to self. Further studies are warranted to confirm the expression and immunogenicity of UTR neoantigens.

Leveraging transformers-based language models in proteome bioinformatics

In recent years, the rapid growth of biological data has increased interest in using bioinformatics to analyze and interpret this data. Proteomics, which studies the structure, function, and interactions of proteins, is a crucial area of bioinformatics. Using natural language processing (NLP) techniques in proteomics is an emerging field that combines machine learning and text mining to analyze biological data. Recently, transformer-based NLP models have gained significant attention for their ability to process variable-length input sequences in parallel, using self-attention mechanisms to capture long-range dependencies. In this review paper, we discuss the recent advancements in transformer-based NLP models in proteome bioinformatics and examine their advantages, limitations, and potential applications to improve the accuracy and efficiency of various tasks. Additionally, we highlight the challenges and future directions of using these models in proteome bioinformatics research. Overall, this review provides valuable insights into the potential of transformer-based NLP models to revolutionize proteome bioinformatics.

From intuition to AI: evolution of small molecule representations in drug discovery

Within drug discovery, the goal of AI scientists and cheminformaticians is to help identify molecular starting points that will develop into safe and efficacious drugs while reducing costs, time and failure rates. To achieve this goal, it is crucial to represent molecules in a digital format that makes them machine-readable and facilitates the accurate prediction of properties that drive decision-making. Over the years, molecular representations have evolved from intuitive and human-readable formats to bespoke numerical descriptors and fingerprints, and now to learned representations that capture patterns and salient features across vast chemical spaces. Among these, sequence-based and graph-based representations of small molecules have become highly popular. However, each approach has strengths and weaknesses across dimensions such as generality, computational cost, inversibility for generative applications and interpretability, which can be critical in informing practitioners' decisions. As the drug discovery landscape evolves, opportunities for innovation continue to emerge. These include the creation of molecular representations for high-value, low-data regimes, the distillation of broader biological and chemical knowledge into novel learned representations and the modeling of up-and-coming therapeutic modalities.

What Can Ribo-Seq, Immunopeptidomics, and Proteomics Tell Us About the Noncanonical Proteome?

Ribosome profiling (Ribo-Seq) has proven transformative for our understanding of the human genome and proteome by illuminating thousands of noncanonical sites of ribosome translation outside the currently annotated coding sequences (CDSs). A conservative estimate suggests that at least 7000 noncanonical ORFs are translated, which, at first glance, has the potential to expand the number of human protein CDSs by 30%, from ∼19,500 annotated CDSs to over 26,000 annotated CDSs. Yet, additional scrutiny of these ORFs has raised numerous questions about what fraction of them truly produce a protein product and what fraction of those can be understood as proteins according to conventional understanding of the term. Adding further complication is the fact that published estimates of noncanonical ORFs vary widely by around 30-fold, from several thousand to several hundred thousand. The summation of this research has left the genomics and proteomics communities both excited by the prospect of new coding regions in the human genome but searching for guidance on how to proceed. Here, we discuss the current state of noncanonical ORF research, databases, and interpretation, focusing on how to assess whether a given ORF can be said to be "protein coding."

What can Ribo-seq and proteomics tell us about the non-canonical proteome?

Ribosome profiling (Ribo-seq) has proven transformative for our understanding of the human genome and proteome by illuminating thousands of non-canonical sites of ribosome translation outside of the currently annotated coding sequences (CDSs). A conservative estimate suggests that at least 7,000 non-canonical open reading frames (ORFs) are translated, which, at first glance, has the potential to expand the number of human protein-coding sequences by 30%, from ∼19,500 annotated CDSs to over 26,000. Yet, additional scrutiny of these ORFs has raised numerous questions about what fraction of them truly produce a protein product and what fraction of those can be understood as proteins according to conventional understanding of the term. Adding further complication is the fact that published estimates of non-canonical ORFs vary widely by around 30-fold, from several thousand to several hundred thousand. The summation of this research has left the genomics and proteomics communities both excited by the prospect of new coding regions in the human genome, but searching for guidance on how to proceed. Here, we discuss the current state of non-canonical ORF research, databases, and interpretation, focusing on how to assess whether a given ORF can be said to be "protein-coding".

In brief

The human genome encodes thousands of non-canonical open reading frames (ORFs) in addition to protein-coding genes. As a nascent field, many questions remain regarding non-canonical ORFs. How many exist? Do they encode proteins? What level of evidence is needed for their verification? Central to these debates has been the advent of ribosome profiling (Ribo-seq) as a method to discern genome-wide ribosome occupancy, and immunopeptidomics as a method to detect peptides that are processed and presented by MHC molecules and not observed in traditional proteomics experiments. This article provides a synthesis of the current state of non-canonical ORF research and proposes standards for their future investigation and reporting.

Highlights

Combined use of Ribo-seq and proteomics-based methods enables optimal confidence in detecting non-canonical ORFs and their protein products.Ribo-seq can provide more sensitive detection of non-canonical ORFs, but data quality and analytical pipelines will impact results.Non-canonical ORF catalogs are diverse and span both high-stringency and low-stringency ORF nominations.A framework for standardized non-canonical ORF evidence will advance the research field.

Discovering microproteins: making the most of ribosome profiling data

Building a reference set of protein-coding open reading frames (ORFs) has revolutionized biological process discovery and understanding. Traditionally, gene models have been confirmed using cDNA sequencing and encoded translated regions inferred using sequence-based detection of start and stop combinations longer than 100 amino-acids to prevent false positives. This has led to small ORFs (smORFs) and their encoded proteins left un-annotated. Ribo-seq allows deciphering translated regions from untranslated irrespective of the length. In this review, we describe the power of Ribo-seq data in detection of smORFs while discussing the major challenge posed by data-quality, -depth and -sparseness in identifying the start and end of smORF translation. In particular, we outline smORF cataloguing efforts in humans and the large differences that have arisen due to variation in data, methods and assumptions. Although current versions of smORF reference sets can already be used as a powerful tool for hypothesis generation, we recommend that future editions should consider these data limitations and adopt unified processing for the community to establish a canonical catalogue of translated smORFs.