Comparison of Alternative Splicing Landscapes Revealed by Long-Read Sequencing in Hepatocyte-Derived HepG2 and Huh7 Cultured Cells and Human Liver Tissue.

The long-read RNA sequencing developed by Oxford Nanopore Technologies provides a direct quantification of transcript isoforms, thereby making it possible to present alternative splicing (AS) profiles as arrays of single splice variants with different abundances. Additionally, AS profiles can be presented as arrays of genes characterized by the degree of alternative splicing (the DAS-the number of detected splice variants per gene). Here, we successfully utilized the DAS to reveal biological pathways influenced by the alterations in AS in human liver tissue and the hepatocyte-derived malignant cell lines HepG2 and Huh7, thus employing the mathematical algorithm of gene set enrichment analysis. Furthermore, analysis of the AS profiles as abundances of single splice variants by using the graded tissue specificity index τ provided the selection of the groups of genes expressing particular splice variants specifically in liver tissue, HepG2 cells, and Huh7 cells. The majority of these splice variants were translated into proteins products and appeal to be in focus regarding further insights into the mechanisms underlying cell malignization. The used metrics are intrinsically suitable for transcriptome-wide AS profiling using long-read sequencing.

Principal Component Analysis of Alternative Splicing Profiles Revealed by Long-Read ONT Sequencing in Human Liver Tissue and Hepatocyte-Derived HepG2 and Huh7 Cell Lines.

The long-read RNA sequencing developed by Oxford Nanopore Technology provides a direct quantification of transcript isoforms. That makes the number of transcript isoforms per gene an intrinsically suitable metric for alternative splicing (AS) profiling in the application to this particular type of RNA sequencing. By using this simple metric and recruiting principal component analysis (PCA) as a tool to visualize the high-dimensional transcriptomic data, we were able to group biospecimens of normal human liver tissue and hepatocyte-derived malignant HepG2 and Huh7 cells into clear clusters in a 2D space. For the transcriptome-wide analysis, the clustering was observed regardless whether all genes were included in analysis or only those expressed in all biospecimens tested. However, in the application to a particular set of genes known as pharmacogenes, which are involved in drug metabolism, the clustering worsened dramatically in the latter case. Based on PCA data, the subsets of genes most contributing to biospecimens' grouping into clusters were selected and subjected to gene ontology analysis that allowed us to determine the top 20 biological processes among which translation and processes related to its regulation dominate. The suggested metrics can be a useful addition to the existing metrics for describing AS profiles, especially in application to transcriptome studies with long-read sequencing.

Workability of mRNA Sequencing for Predicting Protein Abundance.

Transcriptomics methods (RNA-Seq, PCR) today are more routine and reproducible than proteomics methods, i.e., both mass spectrometry and immunochemical analysis. For this reason, most scientific studies are limited to assessing the level of mRNA content. At the same time, protein content (and its post-translational status) largely determines the cell's state and behavior. Such a forced extrapolation of conclusions from the transcriptome to the proteome often seems unjustified. The ratios of "transcript-protein" pairs can vary by several orders of magnitude for different genes. As a rule, the correlation coefficient between transcriptome-proteome levels for different tissues does not exceed 0.3-0.5. Several characteristics determine the ratio between the content of mRNA and protein: among them, the rate of movement of the ribosome along the mRNA and the number of free ribosomes in the cell, the availability of tRNA, the secondary structure, and the localization of the transcript. The technical features of the experimental methods also significantly influence the levels of the transcript and protein of the corresponding gene on the outcome of the comparison. Given the above biological features and the performance of experimental and bioinformatic approaches, one may develop various models to predict proteomic profiles based on transcriptomic data. This review is devoted to the ability of RNA sequencing methods for protein abundance prediction.

Linking Clinical Blood Metabogram and Gut Microbiota.

Recently, a clinical blood metabogram was developed as a fast, low-cost and reproducible test that allows the implementation of metabolomics in clinical practice. The components of the metabogram are functionally related groups of blood metabolites associated with humoral regulation, the metabolism of lipids, carbohydrates and amines, lipid intake into the organism, and liver function, thereby providing clinically relevant information. It is known that the gut microbiota affects the blood metabolome, and the components of the blood metabolome may affect the composition of the gut microbiota. Therefore, before using the metabogram in the clinic, the link between the metabogram components and the level of gut microorganisms should be established. For this purpose, the metabogram and microbiota data were obtained in this work for the same individuals. Metabograms of blood plasma were obtained by direct mass spectrometry of blood plasma, and the gut microbiome was determined by a culture-based method and real-time polymerase chain reaction (PCR). This study involved healthy volunteers and individuals with varying degrees of deviation in body weight (n = 44). A correlation analysis determined which metabogram components are linked to which gut microorganisms and the strength of this link. Moreover, diagnostic parameters (sensitivity, specificity and accuracy) confirmed the capacity of metabogram components to be used for diagnosing gut microbiota alterations. Therefore, the obtained results allow the use of the metabogram in a clinical setting, taking into account its relationship with gut microbiota.

Analysis of Single Biomacromolecules and Viruses: Is It a Myth or Reality?

The beginning of the twenty-first century witnessed novel breakthrough research directions in the life sciences, such as genomics, transcriptomics, translatomics, proteomics, metabolomics, and bioinformatics. A newly developed single-molecule approach addresses the physical and chemical properties and the functional activity of single (individual) biomacromolecules and viral particles. Within the alternative approach, the combination of "single-molecule approaches" is opposed to "omics approaches". This new approach is fundamentally unique in terms of its research object (a single biomacromolecule). Most studies are currently performed using postgenomic technologies that allow the properties of several hundreds of millions or even billions of biomacromolecules to be analyzed. This paper discusses the relevance and theoretical, methodological, and practical issues related to the development potential of a single-molecule approach using methods based on molecular detectors.

Blood Plasma Proteome: A Meta-Analysis of the Results of Protein Quantification in Human Blood by Targeted Mass Spectrometry.

A meta-analysis of the results of targeted quantitative screening of human blood plasma was performed to generate a reference standard kit that can be used for health analytics. The panel included 53 of the 296 proteins that form a "stable" part of the proteome of a healthy individual; these proteins were found in at least 70% of samples and were characterized by an interindividual coefficient of variation <40%. The concentration range of the selected proteins was 10−10−10−3 M and enrichment analysis revealed their association with rare familial diseases. The concentration of ceruloplasmin was reduced by approximately three orders of magnitude in patients with neurological disorders compared to healthy volunteers, and those of gelsolin isoform 1 and complement factor H were abruptly reduced in patients with lung adenocarcinoma. Absolute quantitative data of the individual proteome of a healthy and diseased individual can be used as the basis for personalized medicine and health monitoring. Storage over time allows us to identify individual biomarkers in the molecular landscape and prevent pathological conditions.

MS Identification of Blood Plasma Proteins Concentrated on a Photocrosslinker-Modified Surface.

This work demonstrates the use of a modified mica to concentrate proteins, which is required for proteomic profiling of blood plasma by mass spectrometry (MS). The surface of mica substrates, which are routinely used in atomic force microscopy (AFM), was modified with a photocrosslinker to allow "irreversible" binding of proteins via covalent bond formation. This modified substrate was called the AFM chip. This study aimed to determine the role of the surface and crosslinker in the efficient concentration of various types of proteins in plasma over a wide concentration range. The substrate surface was modified with a 4-benzoylbenzoic acid N-succinimidyl ester (SuccBB) photocrosslinker, activated by UV irradiation. AFM chips were incubated with plasma samples from a healthy volunteer at various dilution ratios (102X, 104X, and 106X). Control experiments were performed without UV irradiation to evaluate the contribution of physical protein adsorption to the concentration efficiency. AFM imaging confirmed the presence of protein layers on the chip surface after incubation with the samples. MS analysis of different samples indicated that the proteomic profile of the AFM-visualized layers contained common and unique proteins. In the working series of experiments, 228 proteins were identified on the chip surface for all samples, and 21 proteins were not identified in the control series. In the control series, a total of 220 proteins were identified on the chip surface, seven of which were not found in the working series. In plasma samples at various dilution ratios, a total of 146 proteins were identified without the concentration step, while 17 proteins were not detected in the series using AFM chips. The introduction of a concentration step using AFM chips allowed us to identify more proteins than in plasma samples without this step. We found that AFM chips with a modified surface facilitate the efficient concentration of proteins owing to the adsorption factor and the formation of covalent bonds between the proteins and the chip surface. The results of our study can be applied in the development of highly sensitive analytical systems for determining the complete composition of the plasma proteome.

The Application of Ejaculate-Based Shotgun Proteomics for Male Infertility Screening.

Problems with the male reproductive system are of both medical and social significance. As a rule, spermatozoa and seminal plasma proteomes are investigated separately to assess sperm quality. The current study aimed to compare ejaculate proteomes with spermatozoa and seminal plasma protein profiles regarding the identification of proteins related to fertility scores. A total of 1779, 715, and 2163 proteins were identified in the ejaculate, seminal plasma, and spermatozoa, respectively. Among these datasets, 472 proteins were shared. GO enrichment analysis of the common proteins enabled us to distinguish biological processes such as single fertilization (GO:0007338), spermatid development (GO:0007286), and cell motility (GO:0048870). Among the abundant terms for GO cellular components, zona pellucida receptor complex, sperm fibrous sheath, and outer dense fiber were revealed. Overall, we identified 139 testis-specific proteins. For these proteins, PPI networks that are common in ejaculate, spermatozoa, and seminal plasma were related to the following GO biological processes: cilium movement (GO:0003341), microtubule-based movement (GO:0007018), and sperm motility (GO:0097722). For ejaculate and spermatozoa, they shared 15 common testis-specific proteins with spermatogenesis (GO:0007283) and male gamete generation (GO:0048232). Therefore, we speculated that ejaculate-based proteomics could yield new insights into the peculiar reproductive physiology and spermatozoa function of men and potentially serve as an explanation for male infertility screening.

Comparative proteoinformatics revealed the essentials of SDS impact on HaCaT keratinocytes.

There is no direct evidence supporting that SDS is a carcinogen, so to investigate this fact, we used HaCaT keratinocytes as a model of human epidermal cells. To reveal the candidate proteins and/or pathways characterizing the SDS impact on HaCaT, we proposed comparative proteoinformatics pipeline. For protein extraction, the performance of two sample preparation protocols was assessed: 0.2% SDS-based solubilization combined with the 1DE-gel concentration (Protocol 1) and osmotic shock (Protocol 2). As a result, in SDS-exposed HaCaT cells, Protocol 1 revealed 54 differentially expressed proteins (DEPs) involved in the disease of cellular proliferation (DOID:14566), whereas Protocol 2 found 45 DEPs of the same disease ID. The 'skin cancer' term was a single significant COSMIC term for Protocol 1 DEPs, including those involved in double-strand break repair pathway (BIR, GO:0000727). Considerable upregulation of BIR-associated proteins MCM3, MCM6, and MCM7 was detected. The eightfold increase in MCM6 level was verified by reverse transcription qPCR. Thus, Protocol 1 demonstrated high effectiveness in terms of the total number and sensitivity of MS identifications in HaCaT cell line proteomic analysis. The utility of Protocol 1 was confirmed by the revealed upregulation of cancer-associated MCM6 in HaCaT keratinocytes induced by non-toxic concentration of SDS. Data are available via ProteomeXchange with identifier PXD035202.

Gene-centric coverage of the human liver transcriptome: QPCR, Illumina, and Oxford Nanopore RNA-Seq.

It has been shown that the best coverage of the HepG2 cell line transcriptome encoded by genes of a single chromosome, chromosome 18, is achieved by a combination of two sequencing platforms, Illumina RNA-Seq and Oxford Nanopore Technologies (ONT), using cut-off levels of FPKM > 0 and TPM > 0, respectively. In this study, we investigated the extent to which the combination of these transcriptomic analysis methods makes it possible to achieve a high coverage of the transcriptome encoded by the genes of other human chromosomes. A comparative analysis of transcriptome coverage for various types of biological material was carried out, and the HepG2 cell line transcriptome was compared with the transcriptome of liver tissue cells. In addition, the contribution of variability in the coverage of expressed genes in human transcriptomes to the creation of a draft human transcriptome was evaluated. For human liver tissues, ONT makes an extremely insignificant contribution to the overall coverage of the transcriptome. Thus, to ensure maximum coverage of the liver tissue transcriptome, it is sufficient to apply only one technology: Illumina RNA-Seq (FPKM > 0).

An Open-Source Pipeline for Processing Direct Infusion Mass Spectrometry Data of the Human Plasma Metabolome.

Direct infusion mass spectrometry (DIMS) is growing in popularity as an effective method for the screening of biological samples in clinical metabolomics. Being quick to execute, DIMS generally requires special skills when interpreting the results of measurements. By inspecting the similarities between two-dimensional electrospray ionization with quadrupole time-of-flight (ESI-QTOF) and matrix-assisted laser desorption/ionization (MALDI) mass spectra, the pipeline for processing QTOF mass spectra using open-source packages (MALDIquant, MSnbase and MetaboAnalystR) was tested. Previously, all algorithmic workflows have relied on the application of software either provided by a vendor or privately developed by enthusiasts. Here, we computationally examined two ways of interpreting the DIMS results of human blood metabolomic profiling. The studied spectra were acquired using ESI-QTOF maXis Impact II (Bruker Daltonics, Billerica, MA, USA), then pre-processed using COMPASS/DataAnalysis commercial software and mapped onto the metabolites using in-lab-developed MatLab scripts. Alternatively, in this work we used the open-source packages MALDIquant, for spectrum pre-processing, and MetaboAnalystR, for data interpretation, instead of the low-availability commercial and home-made tools. Using a set of 100 plasma samples (20 from volunteers with normal body mass index and 80 from patients at different stages of obesity), we observed a high degree of concordance in annotated metabolic pathways between the proprietary DataAnalysis/MatLab pipeline and our freely available solution.

Protocol for Increasing the Sensitivity of MS-Based Protein Detection in Human Chorionic Villi.

An important step in the proteomic analysis of missing proteins is the use of a wide range of tissues, optimal extraction, and the processing of protein material in order to ensure the highest sensitivity in downstream protein detection. This work describes a purification protocol for identifying low-abundance proteins in human chorionic villi using the proposed "1DE-gel concentration" method. This involves the removal of SDS in a short electrophoresis run in a stacking gel without protein separation. Following the in-gel digestion of the obtained holistic single protein band, we used the peptide mixture for further LC-MS/MS analysis. Statistically significant results were derived from six datasets, containing three treatments, each from two tissue sources (elective or missed abortions). The 1DE-gel concentration increased the coverage of the chorionic villus proteome. Our approach allowed the identification of 15 low-abundance proteins, of which some had not been previously detected via the mass spectrometry of trophoblasts. In the post hoc data analysis, we found a dubious or uncertain protein (PSG7) encoded on human chromosome 19 according to neXtProt. A proteomic sample preparation workflow with the 1DE-gel concentration can be used as a prospective tool for uncovering the low-abundance part of the human proteome.

Number of Detected Proteins as the Function of the Sensitivity of Proteomic Technology in Human Liver Cells.

AIMS: The main goal of the Russian part of C-HPP is to detect and functionally annotate missing proteins (PE2-PE4) encoded by human chromosome 18. To achieve this goal, it is necessary to use the most sensitive methods of analysis. BACKGROUND: However, identifying such proteins in a complex biological mixture using mass spectrometry (MS)-based methods is difficult due to the insufficient sensitivity of proteomic analysis methods. A possible solution to the problem is the pre-fractionation of a complex biological sample at the sample preparation stage. OBJECTIVE: This study aims to measure the detection limit of SRM SIS analysis using a standard set of UPS1 proteins and find a way to enhance the sensitivity of the analysis and to, detect proteins encoded by the human chromosome 18 in liver tissue samples, and compare the data with transcriptomic analysis of the same samples. METHODS: Mass spectrometry, data-dependent acquisition, selected reaction monitoring, highperformance liquid chromatography, data-dependent acquisition in combination with pre-fractionation by alkaline reversed-phase chromatography, selected reaction monitoring in combination with prefractionation by alkaline reversed-phase chromatography methods were used in this study. RESULTS: The results revealed that 100% of UPS1 proteins in a mixture could only be identified at a concentration of at least 10-9 М. The decrease in concentration leads to protein losses associated with technology sensitivity, and no UPS1 protein is detected at a concentration of 10-13 М. Therefore, the two-dimensional fractionation of samples was applied to improve sensitivity. The human liver tissue was examined by selected reaction monitoring and shotgun methods of MS analysis using onedimensional and two-dimensional fractionation to identify the proteins encoded by human chromosome 18. A total of 134 proteins were identified. The overlap between proteomic and transcriptomic data in human liver tissue was ~50%. CONCLUSION: The sample concentration technique is well suited for a standard UPS1 system that is not contaminated with a complex biological sample. However, it is not suitable for use with a complex biological protein mixture. Thus, it is necessary to develop more sophisticated fractionation systems for the detection of all low-copy proteins. This weak convergence is due to the low sensitivity of proteomic technology compared to transcriptomic approaches. Also, total mRNA was used to perform RNA-seq analysis, but not all detected mRNA molecules could be translated into proteins. This introduces additional uncertainty in the data; in the future, we plan to study only translated mRNA molecules-the translatome. Data is available via ProteomeXchange with identifier PXD026997.

Deep proteomic dataset of human liver samples obtained by two-dimensional sample fractionation coupled with tandem mass spectrometry.

The data was acquired from 3 normal human liver tissues by LC-MS methods. The tissue liver samples from male subjects post mortem were obtained from ILSBio LLC (https://bioivt.com/). Liver tissue was frozen in liquid nitrogen, transported and shipped on dry ice. The proteins were extracted and purified followed up by trypsin hydrolysis. The peptide mixture was aliquoted and analyzed by different LC-MS approaches: one-dimensional shotgun LC-MS, two-dimensional LC-MS, two-dimensional SRM SIS (Selected Reaction Monitoring with Stable Isotope-labeled peptide Standards). The Shotgun assay resulted in a qualitative in-depth human liver proteome, and a semi-quantitative iBAQ (intensity-based absolute quantification) value was calculated to show the relative protein content of the sample. Absolute quantitative concentrations of proteins encoded by human chromosome 18 using SRM SIS were obtained.

Prediction of Monomeric and Dimeric Structures of CYP102A1 Using AlphaFold2 and AlphaFold Multimer and Assessment of Point Mutation Effect on the Efficiency of Intra- and Interprotein Electron Transfer.

The three-dimensional structure of monomers and homodimers of CYP102A1/WT (wild-type) proteins and their A83F and A83I mutant forms was predicted using the AlphaFold2 (AF2) and AlphaFold Multimer (AFMultimer) programs, which were compared with the rate constants of hydroxylation reactions of these enzyme forms to determine the efficiency of intra- and interprotein electron transport in the CYP102A1 hydroxylase system. The electron transfer rate constants (ket), which determine the rate of indole hydroxylation by the CYP102A1 system, were calculated based on the distances (R) between donor-acceptor prosthetic groups (PG) FAD→FMN→HEME of these proteins using factor ß, which describes an exponential decay from R the speed of electron transport (ET) according to the tunnelling mechanism. It was shown that the structure of monomers in the homodimer, calculated using the AlpfaFold Multimer program, is in good agreement with the experimental structures of globular domains (HEME-, FMN-, and FAD-domains) in CYP102A1/WT obtained by X-ray structural analysis, and the structure of isolated monomers predicted in AF2 does not coincide with the structure of monomers in the homodimer, although a high level of similarity in individual domains remains. The structures of monomers and homodimers of A83F and A83I mutants were also calculated, and their structures were compared with the wild-type protein. Significant differences in the structure of all isolated monomers with respect to the structures of monomers in homodimers were also found for them, and at the same time, insignificant differences were revealed for all homodimers. Comparative analysis for CYP102A1/WT between the calculated intra- and interprotein distances FAD→FMN→HEME and the rate constants of hydroxylation in these proteins showed that the distance between prosthetic groups both in the monomer and in the dimer allows the implementation of electron transfer between PGs, which is consistent with experimental literature data about kcat. For the mutant form of monomer A83I, an increase in the distance between PGs was obtained, which can restrict electron transportation compared to WT; however, for the dimer of this protein, a decrease in the distance between PGs was observed compared to the WT form, which can lead to an increase in the electron transfer rate constant and, accordingly, kcat. For the monomer and homodimer of the A83F mutant, the calculations showed an increase in the distance between the PGs compared to the WT form, which should have led to a decrease in the electron transfer rate, but at the same time, for the homodimer, the approach of the aromatic group F262 with heme can speed up transportation for this form and, accordingly, the rate of hydroxylation.

Comparative Metabolomic Study of Drosophila Species with Different Lifespans.

The increase in life expectancy, leading to a rise in the proportion of older people, is accompanied by a prevalence of age-related disorders among the world population, the fight against which today is one of the leading biomedical challenges. Exploring the biological insights concerning the lifespan is one of the ways to provide a background for designing an effective treatment for the increase in healthy years of life. Untargeted direct injection mass spectrometry-based metabolite profiling of 12 species of Drosophila with significant variations in natural lifespans was conducted in this research. A cross-comparison study of metabolomic profiles revealed lifespan signatures of flies. These signatures indicate that lifespan extension is associated with the upregulation of amino acids, phospholipids, and carbohydrate metabolism. Such information provides a metabolome-level view on longevity and may provide a molecular measure of organism age in age-related studies.

Genome of the Single Human Chromosome 18 as a "Gold Standard" for Its Transcriptome.

The cutoff level applied in sequencing analysis varies according to the sequencing technology, sample type, and study purpose, which can largely affect the coverage and reliability of the data obtained. In this study, we aimed to determine the optimal combination of parameters for reliable RNA transcriptome data analysis. Toward this end, we compared the results obtained from different transcriptome analysis platforms (quantitative polymerase chain reaction, Illumina RNASeq, and Oxford Nanopore Technologies MinION) for the transcriptome encoded by human chromosome 18 (Chr 18) using the same sample types (HepG2 cells and liver tissue). A total of 275 protein-coding genes encoded by Chr 18 was taken as the gene set for evaluation. The combination of Illumina RNASeq and MinION nanopore technologies enabled the detection of at least one transcript for each protein-coding gene encoded by Chr 18. This combination also reduced the probability of false-positive detection of low-copy transcripts due to the simultaneous confirmation of the presence of a transcript by the two fundamentally different technologies: short reads essential for reliable detection (Illumina RNASeq) and long-read sequencing data (MinION). The combination of these technologies achieved complete coverage of all 275 protein-coding genes on Chr 18, identifying transcripts with non-zero expression levels. This approach can improve distinguishing the biological and technical reasons for the absence of mRNA detection for a given gene in transcriptomics.

Human Chr18 transcriptome dataset combined from the Illumina HiSeq, ONT MinION, and qPCR data.

The chromosome-centric dataset was created by applying several technologies of transcriptome profiling. The described dataset is available at NCBI repository (BioProject ID PRJNA635536). The dataset referred to the same type of tissue, cell lines, transcriptome sequencing technologies, and was accomplished in a period of 8 years (the first data were obtained in 2013 while the last ones - in 2020). The high-throughput sequencing technologies were employed along with the quantitative PCR (qPCR) approach, for data generation using the gene expression level assessment. qPCR was performed for a limited group of genes, encoded on human chromosome 18, for the Russian part of the Chromosome-Centric Human Proteome Project. The data of high-throughput sequencing are provided as Excel spreadsheets, where the data on FPKM and TMP values were evaluated for the whole transcriptome with both Illumina HiSeq and Oxford Nanopore Technologies MinION sequencing.

ScanBious: Survey for Obesity Genes Using PubMed Abstracts and DisGeNET.

We used automatic text-mining of PubMed abstracts of papers related to obesity, with the aim of revealing that the information used in abstracts reflects the current understanding and key concepts of this widely explored problem. We compared expert data from DisGeNET to the results of an automated MeSH (Medical Subject Heading) search, which was performed by the ScanBious web tool. The analysis provided an overview of the obesity field, highlighting major trends such as physiological conditions, age, and diet, as well as key well-studied genes, such as adiponectin and its receptor. By intersecting the DisGeNET knowledge with the ScanBious results, we deciphered four clusters of obesity-related genes. An initial set of 100+ thousand abstracts and 622 genes was reduced to 19 genes, distributed among just a few groups: heredity, inflammation, intercellular signaling, and cancer. Rapid profiling of articles could drive personalized medicine: if the disease signs of a particular person were superimposed on a general network, then it would be possible to understand which are non-specific (observed in cohorts and, therefore, most likely have known treatment solutions) and which are less investigated, and probably represent a personalized case.

Does Proteomic Mirror Reflect Clinical Characteristics of Obesity?

Obesity is a frightening chronic disease, which has tripled since 1975. It is not expected to slow down staying one of the leading cases of preventable death and resulting in an increased clinical and economic burden. Poor lifestyle choices and excessive intake of "cheap calories" are major contributors to obesity, triggering type 2 diabetes, cardiovascular diseases, and other comorbidities. Understanding the molecular mechanisms responsible for development of obesity is essential as it might result in the introducing of anti-obesity targets and early-stage obesity biomarkers, allowing the distinction between metabolic syndromes. The complex nature of this disease, coupled with the phenomenon of metabolically healthy obesity, inspired us to perform data-centric, hypothesis-generating pilot research, aimed to find correlations between parameters of classic clinical blood tests and proteomic profiles of 104 lean and obese subjects. As the result, we assembled patterns of proteins, which presence or absence allows predicting the weight of the patient fairly well. We believe that such proteomic patterns with high prediction power should facilitate the translation of potential candidates into biomarkers of clinical use for early-stage stratification of obesity therapy.