Protein Identification Algorithms Research Articles

Comprehensive Evaluation of Advanced Imputation Methods for Proteomic Data Acquired via the Label-Free Approach

Mass-spectrometry-based proteomics frequently utilizes label-free quantification strategies due to their cost-effectiveness, methodological simplicity, and capability to identify large numbers of proteins within a single analytical run. Despite these advantages, the prevalence of missing values (MV), which can impact up to 50% of the data matrix, poses a significant challenge by reducing the accuracy, reproducibility, and interpretability of the results. Consequently, effective handling of missing values is crucial for reliable quantitative analysis in proteomic studies. This study systematically evaluated the performance of selected imputation methods for addressing missing values in proteomic dataset. Two protein identification algorithms, FragPipe and MaxQuant, were employed to generate datasets, enabling an assessment of their influence on im-putation efficacy. Ten imputation methods, representing three methodological categories—single-value (LOD, ND, SampMin), local-similarity (kNN, LLS, RF), and global-similarity approaches (LSA, BPCA, PPCA, SVD)—were analyzed. The study also investigated the impact of data logarithmization on imputation performance. The evaluation process was conducted in two stages. First, performance metrics including normalized root mean square error (NRMSE) and the area under the receiver operating characteristic (ROC) curve (AUC) were applied to datasets with artificially introduced missing values. The datasets were designed to mimic varying MV rates (10%, 25%, 50%) and proportions of values missing not at random (MNAR) (0%, 20%, 40%, 80%, 100%). This step enabled the assessment of data characteristics on the relative effectiveness of the imputation methods. Second, the imputation strategies were applied to real proteomic datasets containing natural missing values, focusing on the true-positive (TP) classification of proteins to evaluate their practical utility. The findings highlight that local-similarity-based methods, particularly random forest (RF) and local least-squares (LLS), consistently exhibit robust performance across varying MV scenarios. Furthermore, data logarithmization significantly enhances the effectiveness of global-similarity methods, suggesting it as a beneficial preprocessing step prior to imputation. The study underscores the importance of tailoring imputation strategies to the specific characteristics of the data to maximize the reliability of label-free quantitative proteomics. Interestingly, while the choice of protein identification algorithm (FragPipe vs. MaxQuant) had minimal influence on the overall imputation error, differences in the number of proteins classified as true positives revealed more nuanced effects, emphasizing the interplay between imputation strategies and downstream analysis outcomes. These findings provide a comprehensive framework for improving the accuracy and reproducibility of proteomic analyses through an informed selection of imputation approaches.

Graph-Based Cancer Protein Identification Algorithm Reveals Key Proteins in PPI Networks

Graph-Based Cancer Protein Identification Algorithm Reveals Key Proteins in PPI Networks

A stacking-based algorithm for antifreeze protein identification using combined physicochemical, pseudo amino acid composition, and reduction property features

Antifreeze proteins have wide applications in the medical and food industries. In this study, we propose a stacking-based classifier that can effectively identify antifreeze proteins. Initially, feature extraction was performed in three aspects: reduction properties, scalable pseudo amino acid composition, and physicochemical properties. A hybrid feature set comprised of the combined information from these three categories was obtained. Subsequently, we trained the training set based on LightGBM, XGBoost, and RandomForest algorithms, and the training outcomes were passed to the Logistic algorithm for matching, thereby establishing a stacking algorithm. The proposed algorithm was tested on the test set and an independent validation set. Experimental data indicates that the algorithm achieved a recognition accuracy of 98.3 %, and an accuracy of 98.5 % on the validation set. Lastly, we analyzed the reasons why numerical features achieved high recognition capabilities from multiple aspects. Data dimensionality reduction and the analysis from two-dimensional and three-dimensional views revealed separability between positive and negative samples, and the protein three-dimensional structure further demonstrated significant differences in related features between the two samples. Analysis of the classifier revealed that Hr*Hr, HrHr, and Sc-PseAAC_1, 188D(152,116,57,183) were among the seven most important numerical features affecting algorithm recognition. For Hr*Hr and HrHr, supportive sequence level evidence for the reduction dictionary was found in terms of conservation area analysis, multiple sequence alignment, and amino acid conservative substitution. Moreover, the importance of the reduction dictionary was recognized through a comparative analysis of importance before and after the reduction, realizing the effectiveness of the dictionary in improving feature importance. A decision tree model has been utilized to discern the distinctions between dipeptides associated with the physical and chemical properties of His(H), Iso(I), Leu(L), and Lys(K) and other dipeptides. We finally analyzed the other seven features of importance, and data analysis confirmed that hydrophobicity, secondary structure, charge properties, van der Waals forces, and solvent accessibility are also factors affecting the antifreeze capability of proteins.

Identification of Significant RNA-Binding Proteins in the Process of CD44 Splicing Using the Boosted Beta Regression Algorithm.

The expression of RNA-binding proteins and their interaction with the spliced pre-mRNA are the key factors in determining the final isoform profile. Transmembrane protein CD44 is involved in differentiation, invasion, motility, growth and survival of tumor cells, and is also a commonly accepted marker of cancer stem cells and epithelial-mesenchymal transition. However, the functions of the isoforms of this protein differ significantly. In this paper, we developed a method based on the boosted beta regression algorithm for identification of the significant RNA-binding proteins in the splicing process by modeling the isoform ratio. The application of this method to the analysis of CD44 splicing in colorectal cancer cells revealed 20 significant RNA-binding proteins. Many of them were previously shown as EMT regulators, but for the first time presented as potential CD44 splicing factors.

IDENTIFICATION OF SIGNIFICANT RNA-BINDING PROTEINS IN THE PROCESS OF CD44 SPLICING USING THE BOOSTED BETA REGRESSION ALGORITHM

The expression of RNA-binding proteins and their interaction with the spliced pre-mRNA are the key factors in determining the final isoform profile. Transmembrane protein CD44 is involved in differentiation, invasion, motility, growth and survival of tumor cells, and is also a commonly accepted marker of cancer stem cells and epithelial-mesenchymal transition. However, the functions of the isoforms of this protein differ significantly. In this paper, we developed a method based on the boosted beta regression algorithm for identification of the significant RNA-binding proteins in the splicing process by modeling the isoform ratio. The application of this method to the analysis of CD44 splicing in colorectal cancer cells revealed 20 significant RNA-binding proteins. Many of them were previously shown as EMT regulators, but for the first time presented as potential CD44 splicing factors.

QuoteTarget: A sequence-based transformer protein language model to identify potentially druggable protein targets.

The development of efficient computational methods for drug target protein identification can compensate for the high cost of experiments and is therefore of great significance for drug development. However, existing structure-based drug target protein-identification algorithms are limited by the insufficient number of proteins with experimentally resolved structures. Moreover, sequence-based algorithms cannot effectively extract information from protein sequences and thus display insufficient accuracy. Here, we combined the sequence-based self-supervised pretraining protein language model ESM1b with a graph convolutional neural network classifier to develop an improved, sequence-based drug target protein identification method. This complete model, named QuoteTarget, efficiently encodes proteins based on sequence information alone and achieves an accuracy of 95% with the nonredundant drug target and nondrug target datasets constructed for this study. When applied to all proteins from Homo sapiens, QuoteTarget identified 1213 potential undeveloped drug target proteins. We further inferred residue-binding weights from the well-trained network using the gradient-weighted class activation mapping (Grad-Cam) algorithm. Notably, we found that without any binding site information input, significant residues inferred by the model closely match the experimentally confirmed drug molecule-binding sites. Thus, our work provides a highly effective sequence-based identifier for drug target proteins, as well to yield new insights into recognizing drug molecule-binding sites. The entire model is available at https://github.com/Chenjxjx/drug-target-prediction.

A DISTRIBUTED ALGORITHM FOR PROTEIN IDENTIFICATION FROM TANDEM MASS SPECTROMETRY DATA

Tandem mass spectrometry is an analytical technique widely used in proteomics for the high-throughput characterization of proteins in biological samples. Modern in-depth proteomic studies require the collection of even millions of mass spectra representing short protein fragments (peptides). In order to identify the peptides, the measured spectra are most often scored against a database of amino acid sequences of known proteins. Due to the volume of input data and the sizes of proteomic databases, this is a resource-intensive task, which requires an efficient and scalable computational strategy. Here, we present SparkMS, an algorithm for peptide and protein identification from mass spectrometry data explicitly designed to work in a distributed computational environment. To achieve the required performance and scalability, we use Apache Spark, a modern framework that is becoming increasingly popular not only in the field of “big data” analysis but also in bioinformatics. This paper describes the algorithm in detail and demonstrates its performance on a large proteomic dataset. Experimental results indicate that SparkMS scales with the number of worker nodes and the increasing complexity of the search task. Furthermore, it exhibits a protein identification efficiency comparable to X!Tandem, a widely-used proteomic search engine.

Recent Progress in FD-LC-MS/MS Proteomics Method.

Through the course of our bio-analytical chemistry studies, we developed a novel proteomics analysis method, FD-LC-MS/MS (fluorogenic derivatization-liquid chromatography-tandem mass spectrometry). This method consists of fluorogenic derivatization (FD), LC separation, and detection/quantification of the derivatized proteins, followed by isolation, tryptic digestion of the isolated proteins, and final identification of the isolated proteins using electrospray ionization nano-LC-MS/MS of the generated peptide mixtures with a probability-based protein identification algorithm. In this review, we will present various examples where this method has been used successfully to identify expressed proteins in individual human cells. FD-LC-MS/MS is also suitable for differential proteomics analysis. Here, two biological samples are treated by the same steps mentioned above, and the two chromatograms obtained are compared to identify peaks with different intensities (variation in protein levels). Associated peak fractions are then isolated, and the differentially expressed proteins between the two samples are identified by LC-MS/MS. Several biomarkers for cancers have been identified by FD-LC-MS/MS. For more efficient separation, nano-flow LC with a phenyl-bonded monolithic silica-based capillary column was adopted for cell-expressed intact protein analysis. The derivatized human cell proteins (K562) and yeast cell (Saccharomyces cerevisiae) proteins as model intact cell proteins were analyzed by nano-flow LC with fluorescence detection. More than 1,300 protein peaks were separated/detected from both cells. For straightforward comparison of multiple peak separation profiles, a novel type of chromatogram display, termed the “spiderweb” chromatogram, was developed. A nano-LC-FD-LC-mass spectrometry trial for molecular weight estimation of FD proteins has also been conducted.

Signal Processing for Single Biomolecule Identification Using Nanopores: A Review

Nanopore sensors provide an innovative platform for rapid and label-free biomolecule detection. The analytical information extracted from resistive pulse sensing (RPS) measurement, though powerful has several limitations. To address these challenges, various signal conditioning algorithms have been researched which include efficient clustering and classification of data, applying probabilistic models to reduce information overlap and also leveraging transient analysis of ionic current instead of the steady state RPS measurement. Additionally, as the current is typically in few picoampere regime with high transition rates, its faithful measurement requires precision instrumentation like a patch clamp amplifier with low noise and high bandwidth. However, for practical deployment of such systems, it is necessary to develop portable electronic interface. Significant efforts have been directed globally to develop on-chip customized integrated circuits with low noise, high bandwidth amplifiers. Thus, it is absolutely timely and pertinent to conduct a review on the various signal processing schemes and electronic interfaces for single biomolecule identification using nanopores which will enable to comprehend the current status for its deployment as a commercial platform for genomics and ultrasensitive detection. To the best of our knowledge, the data extraction methodologies and electronic interface schemes for nanopores have not yet been summarized. In this paper, a comprehensive review emphasizing the signal processing algorithms for DNA sequencing and protein identification in complex analyte using functionalized and non-functionalized nanopores has been presented along with the on-chip low noise electronic interfaces required for field deployable measurement.

What Is New in (Plant) Proteomics Methods and Protocols: The 2015-2019 Quinquennium.

The third edition of "Plant Proteomics Methods and Protocols," with the title "Advances in Proteomics Techniques, Data Validation, and Integration with Other Classic and -Omics Approaches in the Systems Biology Direction," was conceived as being based on the success of the previous editions, and the continuous advances and improvements in proteomic techniques, equipment, and bioinformatics tools, and their uses in basic and translational plant biology research that has occurred in the past 5years (in round figures, of around 22,000 publications referenced in WoS, 2000 were devoted to plants).The monograph contains 29 chapters with detailed proteomics protocols commonly employed in plant biology research. They present recent advances at all workflow stages, starting from the laboratory (tissue and cell fractionation, protein extraction, depletion, purification, separation, MS analysis, quantification) and ending on the computer (algorithms for protein identification and quantification, bioinformatics tools for data analysis, databases and repositories). At the end of each chapter there are enough explanatory notes and comments to make the protocols easily applicable to other biological systems and/or studies, discussing limitations, artifacts, or pitfalls. For that reason, as with the previous editions, it would be especially useful for beginners or novices.Out of the 29 chapters, six are devoted to descriptive proteomics, with a special emphasis on subcellular protein profiling (Chapters 5 - 10 ), six to PTMs (Chapters 11 , and 14 - 18 ), three to protein interactions (Chapters 19 - 21 ), and two to specific proteins, peroxidases (Chapter 24 ) and proteases and protease inhibitors (Chapter 26 ). The book reflects the new trajectory in MS-based protein identification and quantification, moving from the classic gel-based approaches to the most recent labeling (Chapters 10 , 11 , 29 ), shotgun (Chapters 5 , 7 , 12 , 15 ), parallel reaction monitoring (Chapter 16 ), and targeted data acquisition (Chapter 13 ). MS imaging (Chapter 25 ), the only in vivo MS-based proteomics strategy, is far from being fully optimized and exploited in plant biology research. A confident protein identification and quantitation, especially in orphan species, of low-abundance proteins, is still a challenging task (Chapters 4 , 28 ).What is really new is the use of different techniques for proteomics data validation and their integration into other classic and -omics approaches in the systems biology direction. Chapter 2 reports on multiple extractions in a single experiment of the different biomolecules, nucleic acids, proteins, and metabolites. Chapter 27 describes how metabolic pathways can be reconstructed from multiple -omics data, and Chapter 3 network building. Finally, Chapters 22 and 23 deal with, respectively, the search for allele-specific proteins and proteogenomics.Around 200 groups were, almost 1year ago, invited to take part in this edition. Unfortunately, only 10% of them kindly accepted. My gratitude to those who accepted our invitation but also to those who did not, as all of them have contributed to the plant proteomics field. I will enlist, in this introductory chapter, following my own judgment, some of the relevant papers published in the past 5years, those that have shown us how to enhance and exploit the potential of proteomics in plant biology research, without aiming at giving a too exhaustive list.

Clinical Pharmaceutical Research Based on New Proteome Analysis Based on Chromatographic Separation

Comprehensive identification of antigens in immune complexes (IC-antigens) is beneficial to provide insights into pathophysiology and could form the basis for novel diagnostic and treatment strategies for many immune-related diseases. Immune complexome analysis is a method for comprehensively identifying and profiling IC-antigens in biological fluids (such as serum and cerebrospinal fluid). We applied this strategy to the analysis of circulating ICs in autoimmune diseases (rheumatoid arthritis, Sjögren's syndrome, systemic scleroderma, and systemic lupus erythematosus), infectious diseases, and cancers. Fluorogenic derivatization-liquid chromatography-tandem mass spectrometry (FD-LC-MS/MS) consists of fluorogenic derivatization of proteins, followed by HPLC of the derivatized proteins, isolation of the proteins differentially expressed in a certain group, enzymatic digestion of the isolated proteins followed by LC-tandem MS using a database-searching algorithm for protein identification. We have applied this method to understand the cardioprotective effect of pre-administration of docetaxel in adriamycin/docetaxel combination anti-cancer therapy, and the cellular processes that are affected by non-steroidal anti-inflammatory drugs (NSAIDs) in mouse stomach tissue during ulcer formation.

A protein identification algorithm for tandem mass spectrometry by incorporating the abundance of mRNA into a binomial probability scoring model

Peptide-spectrum matches (PSM) scoring between the experimental and theoretical spectrum is a key step in the identification of proteins using mass spectrometry (MS)-based proteomics analyses. Efficient protein identification using MS/MS data remains a challenge. The strategy of using RNA-seq data increases the number of proteins identified by re-constructing the custom search database and integrating mRNA abundance into the false discovery rate of post-PSM. However, this process lacks an algorithm that can allow the incorporation of mRNA abundance into the key scoring model of PSM. Therefore, we developed a novel PSM scoring model, which incorporates mRNA abundance for improved peptide and protein identification. In the new algorithm, abundance information of mRNA was transformed to the prior probability of protein identification and integrated to re-score in PSM using the binomial probability distribution model. Compared with other algorithms using five MS/MS datasets, the results showed that the least improvement ratios of peptide and protein groups were 3.39%–9.79% and 0.48%–8.16% in different datasets (human, rat, zebrafish, yeast, and Arabidopsis thaliana). The new strategy offers an effective solution for MS-based identification of peptides and proteins. SignificanceThe new algorithm identifies proteins by quantifying mRNA abundance (FPKM) and incorporating it into a scoring model for peptide-spectrum matches. It is important to improve peptide and protein identification from MS/MS datasets in proteomics research.

Identification of Poly(ethylene glycol) and Poly(ethylene glycol)-Based Detergents Using Peptide Search Engines

Poly(ethylene glycol) (PEG) is one of the most common polymer contaminations in mass spectrometry (MS) samples. At present, the detection of PEG and other polymers relies largely on manual inspection of raw data, which is laborious and frequently difficult due to sample complexity and retention characteristics of polymer species in reversed-phase chromatography. We developed a new strategy for the automated identification of PEG molecules from tandem mass spectrometry (MS/MS) data using protein identification algorithms in combination with a database containing "PEG-proteins". Through definition of variable modifications, we extend the approach for the identification of commonly used PEG-based detergents. We exemplify the identification of different types of polymers by static nanoelectrospray tandem mass spectrometry (nanoESI-MS/MS) analysis of pure detergent solutions and data analysis using Mascot. Analysis of liquid chromatography-tandem mass spectrometry (LC-MS/MS) runs of a PEG-contaminated sample by Mascot identified 806 PEG spectra originating from four PEG species using a defined set of modifications covering PEG and common PEG-based detergents. Further characterization of the sample for unidentified PEG species using error-tolerant and mass-tolerant searches resulted in identification of 3409 and 3187 PEG-related MS/MS spectra, respectively. We further demonstrate the applicability of the strategy for Protein Pilot and MaxQuant.

Quantification of the plant terpenome: predicted versus actual emission potentials

Plant essential oils are complex mixtures of volatile organic compounds, which play indispensable roles in communication, defense, and adaptive evolution. The complete chemical library produced by a plant is referred to as its terpenome. The potential biological information stored in essential oil composition data can provide an insight into the silent language of plants, as well as roles of terpene emissions in direct and indirect defense, and for playing a crucial role in adaptive evolution. In this work, we have attempted to measure the plant terpenome from a global perspective. One way of measuring the terpenome is to observe and record actual emissions in natural conditions, and this has been in practice for over a century through variously evolving methods of comprehensive GC–MS and HPLC. An alternative method is a knowledge-based prediction of the terpenome, and this method has gained popularity in recent years, with the advent of large-scale genome sequencing technologies. Over the past decade, our laboratory has been involved in compilation and investigation of the plant terpenome using both these methods and this has offered us the opportunity to compare and contrast data from actual and potential emissions, in order to better understand the terpenome and its roles in primary, secondary and adaptive metabolism. We have used emission data in conjunction with genomic data in order to understand how a plant creates the so-called final terpenome, specific to itself, and whether or not plants tap the complete potential for terpene biosynthesis at their disposal according to their genomes. For measuring actual emissions, we have used EssOilDB (the ESSential OIL DataBase), the largest contextual web resource for phytochemicals and for measuring the total plant potential for emissions, we have used TERZYME, an automated algorithm for identification and analysis of genes and proteins involved in isoprenoid biosynthesis.

IQuant: an automated pipeline for quantitative proteomics based upon isobaric tags.

Quantitative proteomics technology based on isobaric tags is playing an important role in proteomic investigation. In this paper, we present an automated software, named IQuant, which integrates a postprocessing tool of protein identification and advanced statistical algorithms to process the MS/MS signals generated from the peptides labeled by isobaric tags and aims at proteomics quantification. The software of IQuant, which is freely downloaded at http://sourceforge.net/projects/iquant/, can run from a graphical user interface and a command-line interface, and can work on both Windows and Linux systems.

RTANDEM, an R/Bioconductor package for MS/MS protein identification

rTANDEM is an R/Bioconductor package that interfaces the X!Tandem protein identification algorithm. The package can run the multi-threaded algorithm on proteomic data files directly from R. It also provides functions to convert search parameters and results to/from R as well as functions to manipulate parameters and automate searches. An associated R package, shinyTANDEM, provides a web-based graphical interface to visualize and interpret the results. Together, those two packages form an entry point for a general MS/MS-based proteomic pipeline in R/Bioconductor. rTANDEM and shinyTANDEM are distributed in R/Bioconductor, http://bioconductor.org/packages/release/bioc/. The packages are under open licenses (GPL-3 and Artistice-1.0). frederic.fournier@crchuq.ulaval.ca or arnaud.droit@crchuq.ulaval.ca Supplementary data are available at Bioinformatics online.

A Novel Essential Protein Identification Algorithm Based on the Integration of Local Network Topology and Gene Ontology

A Novel Essential Protein Identification Algorithm Based on the Integration of Local Network Topology and Gene Ontology

Mitochondria proteome profiling: A comparative analysis between gel- and gel-free approaches

Mitochondrial proteomics emerged aiming to disclose the dynamics of mitochondria under various pathophysiological conditions. In the present study we investigated the relative merits of gel-based (2DE and SDS-LC) and gel-free (2D-LC) protein separation approaches and protein identification algorithms (Mascot and Paragon) in the proteome profiling of mitochondria isolated from cultured fibroblasts, a sample traditionally used for diagnosis purposes.Combining data retrieved from 2DE, 2D-LC and SDS-LC and search methods, a total of 696 non-redundant proteins were identified. An overlap of only 19% between the proteins identified by the three different methods was observed when Mascot and Paragon were used. Regarding protein ID, a consistency in the number of identified proteins per sample was noticed for 2DE approach. Independent of the methodological approach chosen, it was noticed that the predominance in mitochondria of hydrophilic proteins with 20–50kDa and pI 5–6 and 8–9; however, 2D-LC and SDS-LC allowed the enrichment of proteins with a mass below 30kDa and of basic proteins with pI values above 8. In conclusion, data from the present study highlight the power of integrating different separation technologies and protein identification algorithms.

PILOT_PROTEIN: Identification of Unmodified and Modified Proteins via High-Resolution Mass Spectrometry and Mixed-Integer Linear Optimization

A novel protein identification framework, PILOT_PROTEIN, has been developed to construct a comprehensive list of all unmodified proteins that are present in a living sample. It uses the peptide identification results from the PILOT_SEQUEL algorithm to initially determine all unmodified proteins within the sample. Using a rigorous biclustering approach that groups incorrect peptide sequences with other homologous sequences, the number of false positives reported is minimized. A sequence tag procedure is then incorporated along with the untargeted PTM identification algorithm, PILOT_PTM, to determine a list of all modification types and sites for each protein. The unmodified protein identification algorithm, PILOT_PROTEIN, is compared to the methods SEQUEST, InsPecT, X!Tandem, VEMS, and ProteinProspector using both prepared protein samples and a more complex chromatin digest. The algorithm demonstrates superior protein identification accuracy with a lower false positive rate. All materials are freely available to the scientific community at http://pumpd.princeton.edu.

Effect of iTRAQ Labeling on the Relative Abundance of Peptide Fragment Ions Produced by MALDI-MS/MS

The identification of proteins in proteomics experiments is usually based on mass information derived from tandem mass spectrometry data. To improve the performance of the identification algorithms, additional information available in the fragment peak intensity patterns has been shown to be useful. In this study, we consider the effect of iTRAQ labeling on the fragment peak intensity patterns of singly charged peptides from MALDI tandem MS data. The presence of an iTRAQ-modified basic group on the N-terminus leads to a more pronounced set of b-ion peaks and distinct changes in the abundance of specific peptide types. We performed a simple intensity prediction by using a decision-tree machine learning approach and were able to show that the relative ion abundance in a spectrum can be correctly predicted and distinguished from closely related sequences. This information will be useful for the development of improved method-specific intensity-based protein identification algorithms.