Immunopeptidomics-based identification of naturally presented non-canonical circRNA-derived peptides

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: March 15, 2024

Circular RNAs (circRNAs) are covalently closed non-coding lacking the 5' cap and poly-A tail. Nevertheless, it has been demonstrated that certain circRNAs can undergo active translation. Therefore, aberrantly expressed in human cancers could be an unexplored source of tumor-specific antigens, potentially mediating anti-tumor T cell responses. This study presents immunopeptidomics workflow with a specific focus on generating circRNA-specific protein fasta reference. The main goal this is to streamline process identifying validating leukocyte antigen (HLA) bound peptides originating from circRNAs. We increase analytical stringency our by retaining identified independently two mass spectrometry search engines and/or applying group-specific FDR for canonical-derived circRNA-derived peptides. A subset specifically encoded region spanning back-splice junction (BSJ) validated targeted MS, direct Sanger sequencing respective transcripts. Our identifies 54 unique BSJ-spanning immunopeptidome melanoma lung cancer samples. approach enlarges catalog proteins explored immunotherapy.

Language: Английский

---

Fragment ion intensity prediction improves the identification rate of non-tryptic peptides in timsTOF

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: May 10, 2024

Abstract Immunopeptidomics is crucial for immunotherapy and vaccine development. Because the generation of immunopeptides from their parent proteins does not adhere to clear-cut rules, rather than being able use known digestion patterns, every possible protein subsequence within human leukocyte antigen (HLA) class-specific length restrictions needs be considered during sequence database searching. This leads an inflation search space results in lower spectrum annotation rates. Peptide-spectrum match (PSM) rescoring a powerful enhancement standard searching that boosts performance. We analyze 302,105 unique synthesized non-tryptic peptides ProteomeTools project on timsTOF-Pro generate ground-truth dataset containing 93,227 MS/MS spectra 74,847 peptides, used fine-tune deep learning-based fragment ion intensity prediction model Prosit. demonstrate up 3-fold improvement identification immunopeptides, as well increased detection low input samples.

Language: Английский

---

Metagenome-informed metaproteomics of the human gut microbiome, host, and dietary exposome uncovers signatures of health and inflammatory bowel disease

Cell, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

Language: Английский

---

TIMS²Rescore: A Data Dependent Acquisition-Parallel Accumulation and Serial Fragmentation-Optimized Data-Driven Rescoring Pipeline Based on MS²Rescore

Journal of Proteome Research, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 6, 2025

The high throughput analysis of proteins with mass spectrometry (MS) is highly valuable for understanding human biology, discovering disease biomarkers, identifying therapeutic targets, and exploring pathogen interactions. To achieve these goals, specialized proteomics subfields, including plasma proteomics, immunopeptidomics, metaproteomics, must tackle specific analytical challenges, such as an increased identification ambiguity compared to routine experiments. Technical advancements in MS instrumentation can mitigate issues by acquiring more discerning information at higher sensitivity levels. This exemplified the incorporation ion mobility parallel accumulation serial fragmentation (PASEF) technologies timsTOF instruments. In addition, AI-based bioinformatics solutions help overcome integrating data into workflow. Here, we introduce TIMS2Rescore, a data-driven rescoring workflow optimized DDA-PASEF from platform includes new MS2PIP spectrum prediction models IM2Deep, deep learning-based peptide predictor. Furthermore, fully streamline throughput, TIMS2Rescore directly accepts Bruker raw search results ProteoScape many other engines, Sage PEAKS. We showcase performance on immunopeptidomics (HLA class I II), metaproteomics sets. open-source freely available https://github.com/compomics/tims2rescore.

Language: Английский

---

MS²Rescore 3.0 Is a Modular, Flexible, and User-Friendly Platform to Boost Peptide Identifications, as Showcased with MS Amanda 3.0

Journal of Proteome Research, Journal Year: 2024, Volume and Issue: 23(8), P. 3200 - 3207

Published: March 16, 2024

Rescoring of peptide-spectrum matches (PSMs) has emerged as a standard procedure for the analysis tandem mass spectrometry data. This emphasizes need software maintenance and continuous improvement such algorithms. We introduce MS

Language: Английский

---

Assessment of false discovery rate control in tandem mass spectrometry analysis using entrapment

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 3, 2024

Abstract A pressing statistical challenge in the field of mass spectrometry proteomics is how to assess whether a given software tool provides accurate error control. Each for searching such data uses its own internally implemented methodology reporting and controlling error. Many these tools are closed source, with incompletely documented methodology, strategies validating inconsistent across tools. In this work, we identify three different methods false discovery rate (FDR) control use field, one which invalid, can only provide lower bound rather than an upper bound, valid but under-powered. The result that has very poor understanding well doing respect FDR control, particularly analysis data-independent acquisition (DIA) data. We therefore propose new, more powerful method evaluating setting, then employ method, along existing bounding technique, characterize variety popular search find data-dependent (DDA) generally seem at peptide level, whereas none DIA consistently controls level all datasets investigated. Furthermore, problem becomes much worse when latter evaluated protein level. These results may have significant implications various downstream analyses, since proper potential reduce noise lists thereby boost power.

Language: Английский

---

Massively parallel sample preparation for multiplexed single-cell proteomics using nPOP

Nature Protocols, Journal Year: 2024, Volume and Issue: 19(12), P. 3750 - 3776

Published: Aug. 8, 2024

Language: Английский

---

Koina: Democratizing machine learning for proteomics research

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 3, 2024

Abstract Recent developments in machine-learning (ML) and deep-learning (DL) have immense potential for applications proteomics, such as generating spectral libraries, improving peptide identification, optimizing targeted acquisition modes. Although new ML/DL models various properties are frequently published, the rate at which these adopted by community is slow, mostly due to technical challenges. We believe that, make better use of state-of-the-art models, more attention should be spent on making easy accessible community. To facilitate this, we developed Koina, an open-source containerized, decentralized online-accessible high-performance prediction service that enables model usage any pipeline. Using widely used FragPipe computational platform example, show how Koina can easily integrated with existing proteomics software tools integrations improve data analysis.

Language: Английский

---

Rescoring Peptide Spectrum Matches: Boosting Proteomics Performance by Integrating Peptide Property Predictors into Peptide Identification

Molecular & Cellular Proteomics, Journal Year: 2024, Volume and Issue: 23(7), P. 100798 - 100798

Published: June 12, 2024

Language: Английский

---

Comparative Analysis of Data‐Driven Rescoring Platforms for Improved Peptide Identification in HeLa Digest Samples

PROTEOMICS, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 2, 2025

ABSTRACT Mass spectrometry is a critical tool to understand complex changes in biological processes. Despite significant advances search engine technology, many spectra remain unassigned. This research evaluates the performance of three rescoring platforms, Oktoberfest, MS 2 Rescore, and inSPIRE, using MaxQuant output. The results indicated substantial increase identifications at peptide level (40%–53%) PSM (64%–67%). However, some peptides were lost due limitations processing posttranslational modifications (PTMs)—with up 75% exhibiting PTMs. Each platform displayed distinct strengths weaknesses. For instance, inSPIRE performed best terms unique peptides, while Rescore better for PSMs higher FDR values. Differences stemmed from different sources: original feature selection, type ion series predicted, retention time predictor, PTMs compatibility. Overall, showed superior ability harness results. Taken all together, platforms clearly outperformed results; however, they demanded additional computation (up 77%) manual adjustments. findings here underline necessity integrating into current proteomics pipelines but also address challenges their implementation optimization. Future integrated may help enhance adoption.

Language: Английский

---