Protein Libraries Research Articles

Cavity-Based Discovery of New Fatty Acid Photodecarboxylases.

Light-dependent fatty acid photodecarboxylases (FAPs) hold significant potential for biotechnology, due to their capability to produce alka(e)nes directly from the corresponding (un)saturated natural fatty acids requiring light as the only reagent. This study expands the family of FAPs through cavity-based enzyme discovery methods. Thirty enzyme candidates with potential photodecarboxylation activity were identified by matching the cavities of four related template structures against the Protein Data Bank's flavoproteins, a library of proteins identified via the Foldseek Search Server, and homology models of sequences resulting from BLAST. Subsequent docking experiments narrowed this library to ten promising enzymes, which were expressed and assessed in vitro, identifying four photodecarboxylases. Out of these enzymes, the GMC oxidoreductase from Coccomyxa sp. Obi (CoFAP) was characterized in detail, which revealed high activity in the decarboxylation reactions of palmitic acid and octanoic acid and a broad pH tolerance (pH 6.5-9.5).

Rapid and Specific Detection of Bacillus cereus Using Phage Protein-Based Lateral Flow Assays.

Rapid and precise diagnostic techniques are essential for identifying foodborne pathogens, including Bacillus cereus (B. cereus), which poses significant challenges to food safety. Traditional detection methods are limited by long incubation times and high costs. In this context, gold nanoparticle (AuNP)-based lateral flow assays (LFAs) are emerging as valuable tools for rapid screening. However, the use of antibodies in LFAs faces challenges, including complex production processes, ethical concerns, or variability. Here, we address these challenges by proposing an innovative approach using bacteriophage-derived proteins for pathogen detection on LFAs. We used the engineered endolysin cell-wall-binding domain (CBD) and distal tail proteins (Dit) from bacteriophages that specifically target B. cereus. The protein-binding properties, essential for the formation of efficient capture and detection biointerfaces in LFAs, were extensively characterized from the microstructural to the LFA device level. Machine-learning models leverage knowledge of the protein sequence to predict advantageous protein orientations on the nitrocellulose membrane and AuNPs. The study of the biointerface binding quantified the degree of attachment of AuNPs to bacteria, providing, for the first time, a microscopic model of the number of AuNPs binding to bacteria. It highlighted the binding of up to one hundred 40 nm AuNPs per bacterium in conditions mimicking LFAs. Eventually, phage proteins were demonstrated as efficient bioreceptors in a straightforward LFA prototype combining the two proteins, providing a rapid colorimetric response within 15 min upon the detection of 105 B. cereus cells. Recombinantly produced phage binding proteins present an opportunity to generate a customizable library of proteins with precise binding capabilities, offering a cost-effective and ethical alternative to antibodies. This study enhances our understanding of phage protein biointerfaces, laying the groundwork for their utilization as efficient bioreceptors in LFAs and rapid point-of-care diagnostic assays, thus potentially strengthening public health measures.

Controllable Autolytic Leaky E. coli Platform for the Recovery of Intracellular Proteins.

Escherichia coli is a commonly used platform for the production of heterologous proteins. Extraction and purification of intracellularly expressed recombinant proteins rely on efficient cell disruption. To facilitate downstream processing, controlled autolytic cells have been designed that lyse automatically to release intracellular proteins when triggered with an internal or external signal. In the cases when a weak promoter has to be adopted to control autolysis, cell lysis and product release progress slowly even in the presence of surfactants or other adjuvants. In this study, we report an improved autolytic E. coli strain controlled by a weak promoter with higher efficiency without the use of any facilitating chemical. Cell lysis was initiated upon arabinose-induced expression of T4 lysozyme with N-terminal fusion of amphipathic cell-penetrating peptides via a flexible peptide linker. Furthermore, genes involved in membrane permeability were individually deleted and screened for leaky phenotypes. Deletion of lpp (encoding Braun's lipoprotein) combined with the autolytic system caused 96% cell lysis in 4 h of induction and released 84% or 67% of mCherry or a super large Cas13a fusion protein (160.8 kDa), respectively, in 10 h of induction. This autolytic leaky strain shows great promise for protein recovery and library screening.

Combinatorial effector targeting (COMET) for transcriptional modulation and locus-specific biochemistry.

Understanding how human gene expression is coordinately regulated by functional units of proteins across the genome remains a major biological goal. Here, we present COMET, a high-throughput screening platform for combinatorial effector targeting for the identification of transcriptional modulators. We generate libraries of combinatorial dCas9-based fusion proteins, containing two to six effector domains, allowing us to systematically investigate more than 110,000 combinations of effector proteins at endogenous human loci for their influence on transcription. Importantly, we keep full proteins or domains intact, maintaining catalytic cores and surfaces for protein-protein interactions. We observe more than 5800 significant hits that modulate transcription, we demonstrate cell type specific transcriptional modulation, and we further investigate epistatic relationships between our effector combinations. We validate unexpected combinations as synergistic or buffering, emphasizing COMET as both a method for transcriptional effector discovery, and as a functional genomics tool for identifying novel domain interactions and directing locus-specific biochemistry.

Human IL-22 receptor-targeted small protein antagonist suppress murine DSS-induced colitis

BackgroundHuman interleukin-22 (IL-22) is known as a “dual function” cytokine that acts as a master regulator to maintain homeostasis, structural integrity of the intestinal epithelial barrier, and shielding against bacterial pathogens. On the other hand, the overexpression of IL-22 is associated with hyper-proliferation and recruitment of pathologic effector cells, leading to tissue damage and chronic inflammation in specific diseases including inflammatory bowel disease (IBD). To study a role of IL-22-mediated signaling axis during intestinal inflammation, we generated a set of small protein blockers of IL-22R1 and verified their inhibitory potential on murine model of colitis.MethodsWe used directed evolution of proteins to identify binders of human IL-22 receptor alpha (IL-22R1), designated as ABR ligands. This approach combines the assembly of a highly complex combinatorial protein library derived from small albumin-binding domain scaffold and selection of promising protein variants using ribosome display followed by large-scale ELISA screening. The binding affinity and specificity of ABR variants were analyzed on transfected HEK293T cells by flow cytometry and LigandTracer. Inhibitory function was further verified by competition ELISA, HEK-Blue IL-22 reporter cells, and murine dextran sulfate sodium (DSS)-induced colitis.ResultsWe demonstrate that ABR specifically recognizes transgenic IL-22R1 expressed on HEK293T cells and IL-22R1 on TNFα/IFNγ-activated HaCaT cells. Moreover, some ABR binders compete with the IL-22 cytokine and function as IL-22R1 antagonists in HEK-Blue IL22 reporter cells. In a murine model of DSS-induced acute intestinal inflammation, daily intraperitoneal administration of the best IL-22R1 antagonist, ABR167, suppressed the development of clinical and histological markers of colitis including prevention of mucosal inflammation and architecture deterioration. In addition, ABR167 reduces the DSS-induced increase in mRNA transcript levels of inflammatory cytokines such as IL-1β, IL-6, IL-10, and IL-17A.ConclusionsWe developed small anti-human IL-22R1 blockers with antagonistic properties that ascertain a substantial role of IL-22-mediated signaling in the development of intestinal inflammation. The developed ABR blockers can be useful as a molecular clue for further IBD drug development.

Biophysical mapping of TREM2-ligand interactions reveals shared surfaces for engagement of multiple Alzheimer's disease ligands.

TREM2 is a signaling receptor expressed on microglia that has emerged as an important drug target for Alzheimer's disease and other neurodegenerative diseases. While a number of TREM2 ligands have been identified, little is known regarding the structural details of how they engage. To better understand this, we created a protein library of 28 different TREM2 variants that could be used to map interactions with various ligands using biolayer interferometry. The variants are located in previously identified putative binding surfaces on TREM2 called the hydrophobic site, basic site, and site 2. We found that mutations to the hydrophobic site ablated binding to apoE4 and TDP-43. Competition binding experiments indicated that apoE4 and oAβ42 share overlapping binding sites on TREM2. In contrast, binding to C1q was disrupted most strongly by mutations to the basic site, including R46, with some mutations to the hydrophobic site also attenuating binding, thus suggesting a broader mediation of binding across the two sites. Supporting this, competition experiments indicated that C1q binding could be blocked by both apoE and oAβ42. TREM2 binding to IL-34 was mediated by the basic site at a surface centering on R76. Competition binding experiments validated the unique site for IL-34, showing little to no competition with either oAβ42 or apoE4. However, competition experiments between C1q and IL34 suggest that the ligands compete for binding at the basic site. Altogether, our results suggest that TREM2 utilizes the hydrophobic site (consisting of CDR1, CDR2, and CDR3) as a common site to engage multiple ligands, and uses distinct basic sites to engage others. Our findings imply that pharmaceutical strategies targeting these surfaces might be effective to modulate TREM2 functions.

Enriching productive mutational paths accelerates enzyme evolution

Darwinian evolution has given rise to all the enzymes that enable life on Earth. Mimicking natural selection, scientists have learned to tailor these biocatalysts through recursive cycles of mutation, selection and amplification, often relying on screening large protein libraries to productively modulate the complex interplay between protein structure, dynamics and function. Here we show that by removing destabilizing mutations at the library design stage and taking advantage of recent advances in gene synthesis, we can accelerate the evolution of a computationally designed enzyme. In only five rounds of evolution, we generated a Kemp eliminase—an enzymatic model system for proton transfer from carbon—that accelerates the proton abstraction step >108-fold over the uncatalyzed reaction. Recombining the resulting variant with a previously evolved Kemp eliminase HG3.17, which exhibits similar activity but differs by 29 substitutions, allowed us to chart the topography of the designer enzyme’s fitness landscape, highlighting that a given protein scaffold can accommodate several, equally viable solutions to a specific catalytic problem.

Placing a Full Protein Library Under Pressure

Placing a Full Protein Library Under Pressure

Safe and effective in vivo delivery of DNA and RNA using proteolipid vehicles

Genetic medicines show promise for treating various diseases, yet clinical success has been limited by tolerability, scalability, and immunogenicity issues of current delivery platforms. To overcome these, we developed a proteolipid vehicle (PLV) by combining features from viral and non-viral approaches. PLVs incorporate fusion-associated small transmembrane (FAST) proteins isolated from fusogenic orthoreoviruses into a well-tolerated lipid formulation, using scalable microfluidic mixing. Screening a FAST protein library, we identified a chimeric FAST protein with enhanced membrane fusion activity that improved gene expression from an optimized lipid formulation. Systemically administered FAST-PLVs showed broad biodistribution and effective mRNA and DNA delivery in mouse and non-human primate models. FAST-PLVs show low immunogenicity and maintain activity upon repeat dosing. Systemic administration of follistatin DNA gene therapy with FAST-PLVs raised circulating follistatin levels and significantly increased muscle mass and grip strength. These results demonstrate the promising potential of FAST-PLVs for redosable gene therapies and genetic medicines.

Characterization and engineering of the type 3 secretion system needle monomer from Salmonella through the construction and screening of a comprehensive mutagenesis library.

Protein production strategies in bacteria are often limited due to the need for cell lysis and complicated purification schemes. To avoid these challenges, researchers have developed bacterial strains capable of secreting heterologous protein products outside the cell, but secretion titers often remain too low for commercial applicability. Improved understanding of the link between secretion system structure and its secretory abilities can help overcome the barrier to engineering higher secretion titers. Here, we investigated this link with the PrgI protein, the monomer of the secretory channel of the type 3 secretion system (T3SS) of Salmonella enterica. Despite detailed knowledge of the PrgI needle's assembly and structure, little is known about how its structure influences its secretory capabilities. To study this, we recently constructed a comprehensive codon mutagenesis library of the PrgI protein utilizing a novel one-pot recombineering approach. We then screened this library for functional T3SS assembly and secretion titer by measuring the secretion of alkaline phosphatase using a high-throughput activity assay. This allowed us to construct a first-of-its-kind secretion fitness landscape to characterize the PrgI needle's mutability at each position as well as the mutations which lead to enhanced T3SS secretion. We discovered new design rules for building a functional T3SS as well as identified hypersecreting mutants. This work can be used to increase understanding of the T3SS's assembly and identify further targets for engineering. This work also provides a blueprint for future efforts to engineer other complex protein assemblies through the construction of fitness landscapes.IMPORTANCEProtein secretion offers a simplified alternative method for protein purification from bacterial hosts. However, the current state-of-the-art methods for protein secretion in bacteria are still hindered by low yields relative to traditional protein purification strategies. Engineers are now seeking strategies to enhance protein secretion titers from bacterial hosts, often through genetic manipulations. In this study, we demonstrate that protein engineering strategies focused on altering the secretion apparatus can be a fruitful avenue toward this goal. Specifically, this study focuses on how changes to the PrgI needle protein from the type 3 secretion system from Salmonella enterica can impact secretion titer. We demonstrate that this complex is amenable to comprehensive mutagenesis studies and that this can yield both PrgI variants with increased secretory capabilities and insight into the normal functioning of the type 3 secretion system.

A batch screening technique for the calculation of chromatographic separability

This paper employs a high-throughput parallel batch (microtiter plate) adsorption screen with sequential salt step increases to rapidly generate protein elution profiles for multiple resins at different pHs using a protein library. The chromatographic set used in this work includes single mode, multimodal anion-exchange (MMA), and multimodal cation-exchange (MMC) resins. The protein library consists of proteins with isoelectric points ranging from 5.1 to 11.4 with varying hydrophobicities as determined by their retention on hydrophobic interaction chromatography. The batch sequential experiments are carried out using one protein at a time with a wide set of resins at multiple pH conditions, thus enabling simple microtiter plate detection. A mathematical formulation is then used to determine the first moment of the distributions from each chromatogram (sequential step elution) generated in the parallel batch experiments. Batch data first moments (expressed in salt concentration) are then compared to results obtained from column linear salt gradient elution, and the techniques are shown to be consistent. In addition, first moment data are used to calculate one-resin separability scores, which are a measure of a resin's ability, at a specified pH, to separate the entire set of proteins in the library from one another. Again, the results from the batch and column experiments are shown to be comparable. The first moment data sets were then employed to calculate the two-resin separability scores, which are a measure of the ability of two resins to synergistically separate the entire set of proteins in the library. Importantly, these results based on the two-resin separability performances derived from the batch and column experiments were again shown to be consistent. This approach for rapidly screening large numbers of chromatographic resins and mobile phase conditions for their elution behavior may prove useful for enabling the rapid discovery of new chromatographic ligands and resins.

Development of a Recombineering System for the Acetogen Eubacterium limosum with Cas9 Counterselection for Markerless Genome Engineering.

Eubacterium limosum is a Clostridial acetogen that efficiently utilizes a wide range of single-carbon substrates and contributes to metabolism of health-associated compounds in the human gut microbiota. These traits have led to interest in developing it as a platform for sustainable CO2-based biofuel production to combat carbon emissions, and for exploring the importance of the microbiota in human health. However, synthetic biology and metabolic engineering in E. limosum have been hindered by the inability to rapidly make precise genomic modifications. Here, we screened a diverse library of recombinase proteins to develop a highly efficient oligonucleotide-based recombineering system based on the viral recombinase RecT. Following optimization, the system is capable of catalyzing ssDNA recombination at an efficiency of up to 2%. Addition of a Cas9 counterselection system eliminated unrecombined cells, with up to 100% of viable cells encoding the desired mutation, enabling creation of genomic point mutations in a scarless and markerless manner. We deployed this system to create a clean knockout of the extracellular polymeric substance (EPS) gene cluster, generating a strain incapable of biofilm formation. This approach is rapid and simple, not requiring laborious homology arm cloning, and can readily be retargeted to almost any genomic locus. This work overcomes a major bottleneck in E. limosum genetic engineering by enabling precise genomic modifications, and provides both a roadmap and associated recombinase plasmid library for developing similar systems in other Clostridia of interest.

Microbial diversity and gene abundance in denitrifying bioreactors: A comparison of the woodchip surface biofilm versus the interior wood matrix.

Excessive amounts of nitrogen (N) and phosphorus (P) can lead to eutrophication in water sources. Woodchip bioreactors have shown success in removing N from agricultural runoff, but less is known regarding P removal. Woodchip bioreactors are subsurface basins filled with woodchips installed downgradient of agricultural land to collect and treat drainage runoff. Microorganisms use the woodchips as a carbon (C) source to transform N in the runoff, with unresolved biological impacts on P. This study aims to explore microbial communities present in the bioreactor and determine whether milling woodchips to probe the microbial communities within them reveals hidden microbial diversities or potential activities. Metagenomic sequencing and bioinformatic analyses were performed on six woodchip samples (i.e., three unmilled and three milled) collected from a 10-year-old woodchip bioreactor treating agricultural tile drainage. All samples had similar DNA purity, yield, quality, and microbial diversity regardless of milling. However, when sequences were aligned against various protein libraries, our results indicated greater relative abundance of denitrification and P transformation proteins on the outside of the woodchips (unmilled), while the interior of woodchips (milled) exhibited more functional gene abundance for carbohydrate breakdown. Thus, it may be important to characterize microbial communities both within woodchips, and on woodchip surfaces, to gain a more holistic understanding of coupled biogeochemical cycles on N, P, and C in woodchip bioreactors. Based on these findings, we advise that future microbial research on woodchips (and potentially other permeable organic materials) examine both the surface biofilm and the interior organic material during initial studies. Once researchers determine where specific proteins or enzymes of interest are most prevalent, subsequent studies may then focus on either one or both aspects, as needed.

Construction of a Membrane Yeast Two-Hybrid Library and Screening of MsPYR1-Like Interacting Proteins in Malus sieversii.

To investigate the biological effects of the ABA receptor pyrabactin resistance 1-like (PYR1-like) in Malus sieversii seeds, the proteins interacting with MsPYR1-like were screened by the membrane yeast two-hybrid library based on the split-ubiquitin system, and to construct the bait vector pBT3-SUC-PYR1 for Malus sieversii cDNA library, which had no self-activating effect on the yeast cells of the pPR3-N membrane yeast two-hybrid library. The library titer assay showed that it could meet the requirements for membrane yeast two-hybrid library screening. After sequencing, GenBank database blast, and yeast rotary validation, 28 candidate proteins interacting with MsPYR1-like were obtained, including ribosomal proteins, late embryogenesis abundant proteins, F-actin-capping proteins, phytochrome-interacting proteins, low-temperature-inducible 65kDa protein-like, senescence-associated, PP2C and SnRK2 family members, and unknown proteins. Gene ontology analysis of the interaction proteins was related to plant hormone response and negative regulation of seed germination, overexpression of MsPYR1-like in Arabidopsis negatively regulates seed germination, and the study of the biological roles of MsPYR1-like interacting proteins lays the foundation for revealing the lifting of seed dormancy in Malus sieversii.

Identification of DPP-IV inhibitory peptides derived from buffalo colostrum: Mining through bioinformatics, in silico and in vitro approaches.

Bioactive peptides derived from foods provide physiological health benefits beyond nutrition. This study focused on profiling small peptide inhibitors against two key serine proteases, dipeptidyl peptidase-IV (DPP-IV) and prolyl oligopeptidase (POP). DPP-IV is a well-known protein involved in diverse pathways regulating inflammation, renal, cardiovascular physiology, and glucose homeostasis. POP is yet another key target protein for neurodegenerative disorders. The study evaluated peptide libraries of buffalo colostrum whey and fat globule membrane proteins derived from pepsin and pepsin-pancreatin digestion through in silico web tools and structure-based analysis by molecular docking and binding free-energy estimation, followed by in vitro assay for DPP-IV inhibition for the lead peptides. The bioinformatic study indicated 49 peptides presented motifs with DPP-IV inhibition while 5 peptides with sequences for POP inhibition. In the molecular docking interactions study, 22 peptides interacted with active site residues of DPP-IV and 3 peptides with that of POP. The synthesized peptides, SFVSEVPEL and LTFQHNF inhibited DPP-IV in vitro with an IC50 of 193.5 μM and 1.782 mM, respectively. The study revealed the key residues for inhibition of DPP-IV and POP thus affirming the DPP-IV inhibitory potential of milk-derived peptides.

A Computational Pipeline for Accurate Prioritization of Protein‐Protein Binding Candidates in High‐Throughput Protein Libraries

AbstractIdentifying the interactome for a protein of interest is challenging due to the large number of possible binders. High‐throughput experimental approaches narrow down possible binding partners but often include false positives. Furthermore, they provide no information about what the binding region is (e.g., the binding epitope). We introduce a novel computational pipeline based on an AlphaFold2 (AF) Competitive Binding Assay (AF‐CBA) to identify proteins that bind a target of interest from a pull‐down experiment and the binding epitope. Our focus is on proteins that bind the Extraterminal (ET) domain of Bromo and Extraterminal domain (BET) proteins, but we also introduce nine additional systems to show transferability to other peptide‐protein systems. We describe a series of limitations to the methodology based on intrinsic deficiencies of AF and AF‐CBA to help users identify scenarios where the approach will be most useful. Given the method‘s speed and accuracy, we anticipate its broad applicability to identify binding epitope regions among potential partners, setting the stage for experimental verification.

A Computational Pipeline for Accurate Prioritization of Protein-Protein Binding Candidates in High-Throughput Protein Libraries.

Identifying the interactome for a protein of interest is challenging due to the large number of possible binders. High-throughput experimental approaches narrow down possible binding partners but often include false positives. Furthermore, they provide no information about what the binding region is (e.g., the binding epitope). We introduce a novel computational pipeline based on an AlphaFold2 (AF) Competitive Binding Assay (AF-CBA) to identify proteins that bind a target of interest from a pull-down experiment and the binding epitope. Our focus is on proteins that bind the Extraterminal (ET) domain of Bromo and Extraterminal domain (BET) proteins, but we also introduce nine additional systems to show transferability to other peptide-protein systems. We describe a series of limitations to the methodology based on intrinsic deficiencies of AF and AF-CBA to help users identify scenarios where the approach will be most useful. Given the method's speed and accuracy, we anticipate its broad applicability to identify binding epitope regions among potential partners, setting the stage for experimental verification.

Mutation of the peptide-regulated transcription factor ComR for amidated peptide specificity and heterologous function in Lactiplantibacillus plantarum WCFS1.

There is a growing interest in the use of probiotic bacteria as biosensors for the detection of disease. However, there is a lack of bacterial receptors developed for specific disease biomarkers. Here, we have investigated the use of the peptide-regulated transcription factor ComR from Streptococcus spp. for specific peptide biomarker detection. ComR exhibits a number of attractive features that are potentially exploitable to create a biomolecular switch for engineered biosensor circuitry within the probiotic organism Lactiplantibacillus plantarum WCFS1. Through iterative design-build-test cycles, we developed a genomically integrated, ComR-based biosensor circuit that allowed WCFS1 to detect low nanomolar concentrations of ComR's cognate peptide XIP. By screening a library of ComR proteins with mutant residues substituted at the K100 position, we identified mutations that increased the specificity of ComR toward an amidated version of its cognate peptide, demonstrating the potential for ComR to detect this important class of biomarker.IMPORTANCEUsing bacteria to detect disease is an exciting possibility under active study. Detecting extracellular peptides with specific amino acid sequences would be particularly useful as these are important markers of health and disease (biomarkers). In this work, we show that a probiotic bacteria (Lactiplantibacillus plantarum) can be genetically engineered to detect specific extracellular peptides using the protein ComR from Streptococcus bacteria. In its natural form, ComR allowed the probiotic bacteria to detect a specific peptide, XIP. We then modified XIP to be more like the peptide biomarkers found in humans and engineered ComR so that it activated with this modified XIP and not the original XIP. This newly engineered ComR also worked in the probiotic bacteria, as expected. This suggests that with additional engineering, ComR might be able to activate with human peptide biomarkers and be used by genetically engineered probiotic bacteria to better detect disease.

Synthesis cost-optimal targeted mutant protein libraries

Protein variant libraries produced by site-directed mutagenesis are a useful tool utilized by protein engineers to explore variants with potentially improved properties, such as activity and stability. These libraries are commonly built by selecting residue positions and alternative beneficial mutations for each position. All possible combinations are then constructed and screened, by incorporating degenerate codons at mutation sites. These degenerate codons often encode additional unwanted amino acids or even STOP codons. Our study aims to take advantage of annealing based recombination of oligonucleotides during synthesis and utilize multiple degenerate codons per mutation site to produce targeted protein libraries devoid of unwanted variants. Toward this goal we created an algorithm to calculate the minimum number of degenerate codons necessary to specify any given amino acid set, and a dynamic programming method that uses this algorithm to optimally partition a DNA target sequence with degeneracies into overlapping oligonucleotides, such that the total cost of synthesis of the target mutant protein library is minimized. Computational experiments show that, for a modest increase in DNA synthesis costs, beneficial variant yields in produced mutant libraries are increased by orders of magnitude, an effect particularly pronounced in large combinatorial libraries.

PT-Finder: A multi-modal neural network approach to target identification

Efficient target identification for bioactive compounds, including novel synthetic analogs, is crucial for accelerating the drug discovery pipeline. However, the process of target identification presents significant challenges and is often expensive, which in turn can hinder the drug discovery efforts. To address these challenges machine learning applications have arisen as a promising approach for predicting the targets for novel chemical compounds. These methods allow the exploration of ligand–target interactions, uncovering of biochemical mechanisms, and the investigation of drug repurposing. Typically, the current target identification tools rely on assessing ligand structural similarities. Herein, a multi-modal neural network model was built using a library of proteins, their respective sequences, and active inhibitors. Subsequent validations showed the model to possess accuracy of 82 % and MPRAUC of 0.80. Leveraging the trained model, we developed PT-Finder (Protein Target Finder), a user-friendly offline application that is capable of predicting the target proteins for hundreds of compounds within a few seconds. This combination of offline operation, speed, and accuracy positions PT-Finder as a powerful tool to accelerate drug discovery workflows. PT-Finder and its source codes have been made freely accessible for download at https://github.com/PT-Finder/PT-Finder.