TYK2 inhibitor, screening method based on fusion of deep learning scoring and traditional molecular docking scoring, and application of TYK2 inhibitor

Abstract

The invention discloses a TYK2 inhibitor and a screening method and application thereof based on fusion of deep learning scoring and traditional molecular docking scoring. According to the method, the deep learning scoring result CNNscore and the empirical score obtained by Watvina are combined, and deep learning docking scoring and empirical docking scoring are multiplied, so that the screening efficiency is improved, the problem that the false positive rate of a traditional screening method is too high is solved, and time and cost are saved. The method is further applied to 12 million compounds, in-vitro activity tests verify that two small molecules with TYK2 inhibitory activity are obtained through screening, and the method has application prospects in prevention or / and treatment of TYK2 mediated diseases. Thus, the method facilitates rapid identification of potentially effective compounds from a bulky library of compounds.

Description

Technical Field

[0001] The present invention relates to the field of biomolecular structure prediction, and in particular to a TYK2 inhibitor and a screening method and application thereof based on the fusion of deep learning scoring and traditional molecular docking scoring. Background Art

[0002] Janus kinases (JAKs) are members of a family of non-receptor tyrosine kinases, including JAK1, JAK2, JAK3, and TYK2. These kinases play key roles in the traditional Type I and Type II immune cytokine receptor signaling pathways, making them potential targets for the treatment of a variety of inflammatory and autoimmune diseases. JAK2 plays an important role in multiple physiological processes such as tumor growth, while TYK2 has become a research hotspot because it can interfere with the signaling pathways of IL-23 and IL-12. These two cytokines are closely related to the expansion and survival of TH17 cells, and are therefore related to the pathogenesis of major autoimmune diseases such as psoriasis and inflammatory bowel disease.

[0003] In recent years, with the advancement of drug design and screening technology, the research and development of inhibitors against JAK2 and TYK2 has made significant progress. For example, Deucravacitinib (N-trideuterated methylpyridazine derivative) was approved by the U.S. Food and Drug Administration (FDA) as the first TYK2 JH2 inhibitor, which can effectively target and stabilize the TYK2 JH2 domain, thereby blocking the activity of TYK2JH1. In addition, Deucravacitinib exhibits high selectivity and low toxicity, which has obvious advantages over other FDA-approved JAK inhibitors.

[0004] Molecular docking, as a structure-based virtual screening method, plays an important role in identifying promising bioactive molecules from molecular databases. However, a major challenge of molecular docking is the high proportion of false positive results, that is, the top-ranked molecular conformations are often incorrect. Previous studies have shown that when docking σ2, more than 98% of the top 100 molecules screened from 1.3 billion molecules have incorrect tautomeric forms. The root of this problem lies in the limitations of traditional molecular docking methods.

[0005] Traditional molecular docking methods usually adopt a search and score framework, in which search algorithms are used to explore potential ligand conformations, while scoring functions (SFs) are used to select the optimal ligand conformation and estimate the protein-ligand (PL) binding strength. However, relying on simplified force fields (FFs) and empirical energy terms for binding energy estimation may compromise accuracy. In addition, over-simplification of search algorithms and scoring functions may lead to inaccurate predictions of binding conformations and affinities, ultimately reducing the enrichment capacity of virtual screening (VS). These problems directly lead to a high proportion of false positive results, that is, the top-ranked molecular conformations are often incorrect. Therefore, a strategy that can reduce the false positive rate in virtual screening and improve screening efficiency is particularly important. Summary of the invention

[0006] In order to solve the problems existing in the prior art, the present invention proposes a TYK2 inhibitor and a screening method and application thereof based on the fusion of deep learning scoring and traditional molecular docking scoring. The screening method is highly efficient, and the TYK2 inhibitor obtained by screening has a novel structure and can effectively inhibit the enzymatic activity of TYK2 in vitro.

[0007] In order to achieve the above-mentioned invention object, the present invention adopts the following technical solutions:

[0008] The present invention provides a method for screening TYK2 inhibitors based on the fusion of deep learning scoring and traditional molecular docking scoring, which comprises the following steps:

[0009] S1: Obtain the TYK2 protein structure and small molecules in the compound library, and pre-process them;

[0010] S2: performing molecular docking on the small molecule pretreated in step S1 to obtain the docking conformation and its corresponding Watvina docking score;

[0011] S3: Calculating the relative binding free energy of the molecular docking conformations that meet the Watvina docking score requirements in step S2;

[0012] S4: re-scoring the molecular docking conformations whose relative binding free energy meets the conditions in step S3 based on deep learning, and outputting the deep learning scores;

[0013] S5: performing fusion scoring on the molecules that meet the conditions and are re-scored in step S4, and selecting candidate compounds according to the fusion score ranking;

[0014] S6: Conducting an in vitro activity test on the candidate compounds in step S5 to screen out TYK2 inhibitors with inhibitory activity.

[0015] Furthermore, in step S1, the pretreatment of the protein structure includes deleting redundant chains, deleting solvent molecules, and adding hydrogen atoms and charges.

[0016] Furthermore, when performing molecular docking in step S2, the molecular docking is expanded in three coordinate directions according to the position information of the crystallized ligand of PDB ID 6nzp. Get the 3D information of the docking box.

[0017] Furthermore, in step S3, the small molecules that meet the Watvina docking score requirement are those with a score no higher than -7.

[0018] Furthermore, in step S4, the molecules that meet the relative binding free energy requirement are small molecules with a relative binding energy less than -60 Kcal / mol.

[0019] Furthermore, in step S5, the re-scored molecules that meet the conditions are small molecules with a re-scored CNN score of not less than 0.95.

[0020] Furthermore, the specific steps of the relative binding free energy in step S3 are:

[0021] S31: Use PDBFixer to process the protein structure file pre-processed in step S1, including repairing missing residues, replacing non-standard residues, adding missing atoms and hydrogen atoms; use the Modeller module of OpenMM to add an explicit water model; use openbable to convert the small molecule docking conformation file obtained in step S2 into sdf format; merge the processed protein structure and small molecule, and remove any residues with a distance less than water molecules, avoiding the presence of water molecules that collide with small molecules;

[0022] S32: using LangevinIntegrator of OpenMM to perform energy minimization on the complex system of step S32 to achieve a stable energy state;

[0023] S33: Obtain the potential energy of the complex system of step S32 through OpenMM; construct systems containing only the protein and the ligand respectively, and then calculate their respective potential energies; and obtain the relative binding energy by subtracting the sum of the total energies of the protein and the ligand when they are independent from the total energy of the complex system.

[0024] Furthermore, in step S2, the molecular docking score is Watvina; in step S4, the re-scoring based on deep learning is CNNscore.

[0025] Furthermore, in step S5, the fusion score is obtained by multiplying the Watvina docking score of step S2 by the deep learning score of step S4.

[0026] The present invention also provides a TYK2 inhibitor, which is obtained by screening through the method described above.

[0027] Furthermore, the chemical structure of the TYK2 inhibitor is specifically:

[0028]

[0029] The present invention also provides the use of the TYK2 inhibitor in preparing medicines for treating diseases mediated by TYK2.

[0030] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0031] 1. In order to reduce the false positive rate in virtual screening, the present invention adopts a novel data fusion strategy, which combines the deep learning scoring function (CNNscore) with the empirical scoring function (Watvina). The CNNscore scoring function can learn complex features and evaluate the quality of small molecule and protein binding conformations. Existing studies have shown that CNNscore is superior to the Autodock Vina scoring function in predicting conformations and virtual screening.

[0032] Watvina is a molecular docking software used in the present invention. Based on the molecular docking engine of AutodockVina, Watvina is optimized in terms of scoring function and conformational search algorithm. The scoring mechanism of Watvina is based on empirical factors such as van der Waals forces, hydrogen bonds, polar-polar repulsion, and hydrophobic attraction. Unlike Autodock Vina, Watvina takes into account the contribution of all hydrogen atoms. The conformational search adopts a simplified genetic algorithm-BFGS combination strategy; in addition, the torsion penalty of conjugated rotatable single bonds is also calculated.

[0033] The present invention utilizes data fusion technology to reduce the false positive rate in virtual screening and improve screening efficiency by multiplying the deep learning docking score CNNscore with the empirical docking Watvina score. This fusion strategy significantly improves the accuracy of screening JAK2 and TYK2 inhibitors, especially in screening TYK2 JH2 inhibitors, and helps to quickly identify potential effective compounds from a large compound library.

[0034] 2. The present invention uses layer-by-layer virtual screening filtering and combined with artificial visual detection and analysis, and conducts in vitro activity tests on the selected compounds. The results show that 2 of the 40 compounds screened have good TYK2 inhibitory activity. This proves that the virtual screening method constructed by the present invention is accurate and reliable, and can quickly and efficiently obtain potential inhibitors from a complex compound library, thereby improving screening efficiency and saving experimental costs and time; and further through subsequent in vitro activity experiments, the virtual screening results are further verified for biological activity, which improves the accuracy of the screening results and reduces the false positive rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The results are based on the DUDE dataset test to compare the performance of data fusion technology and Watvina screening alone;

[0036] Figure 2 This is the methodological validation result based on the TYK2 target pair data fusion technology;

[0037] Figure 3 It is to screen the chemical structures of the top 40 small molecules;

[0038] Figure 4 It is the IC50 curve of compounds STL250599 and STK595738. Specific implementation methods

[0040] The technical solution of the present invention is further described in detail with reference to the following specific examples.

[0041] In the following examples, unless otherwise specified, the experimental methods used are all conventional methods, and the materials and reagents used can be purchased from biological or chemical reagent companies.

[0042] Based on the test of DUDE data set, the present invention explores the method of obtaining the best fusion deep learning scoring and traditional molecular docking scoring, and explores the screening performance of the data fusion method on DUDE data set. In addition, the effectiveness of the method in virtual screening is tested on PCBA data set and CASF data set, and finally the effectiveness of the data fusion method in screening TYK2 inhibitors is verified for the TYK2 target. The 12 million super large compound library is virtually screened using Watvina scoring, CNNscore scoring, and MMGBSA calculation, and the data fusion technology is added to the screening process to improve the hit rate of virtual screening. Finally, the compounds are selected for in vitro activity verification, and the lead compounds with TYK2 inhibitory activity are screened.

[0043] Example 1: Exploration and verification of data fusion method

[0044] 1. Research on data fusion methods on DUD-E dataset

[0045] (1) Preparation of DUD-E dataset: 102 protein targets and their corresponding active molecules and decoy molecules were obtained from the DUD-E website (https: / / dude.docking.org / ). Proteins and small molecules were optimized and format converted using the rdkit2pdbqt.py script.

[0046] (2) Molecular docking using Watvina software: Molecular docking of the optimized molecules was performed using Watvina software (https: / / github.com / biocheming / watvina) to obtain the docked conformation and docking score.

[0047] (3) Rescoring using CNNscore: CNNscore (https: / / github.com / gnina / gnina) was used to rescore the conformations after Watvina docking.

[0048] (4) Explore the best data fusion method:

[0049] Method 1: Select the conformation with the lowest Watvina docking score and multiply its score by the corresponding CNNscore.

[0050] Method 2: Select the conformation with the highest CNNscore and multiply its score by the corresponding Watvina docking score.

[0051] Method 3: Calculate the multiplication of the Watvina score and the CNN score for each conformation, and select the conformation with the lowest multiplication score.

[0052] The advantages and disadvantages of the three methods were evaluated by enrichment factor. The results are shown in Table 1. Finally, method 1 was found to be the best data fusion method.

[0053] Table 1 Research on data fusion methods based on DUDE dataset

[0054] Data fusion method EF0.5% EF1% EF5% Method 1 13.40 10.99 5.12 Method 2 10.86 8.87 4.42 Method 3 9.87 8.66 4.75

[0055] 2. Comparison of the screening performance of the data fusion method and Watvina scoring alone on the DUD-E dataset

[0056] The fusion score was obtained based on the minimum docking score of Watvina obtained above and the best data fusion method (method 1), and the top 1%, 5%, and 10% enrichment factors were calculated. Statistical analysis was performed using a two-tailed Mann-Whitney U test to evaluate the difference in screening performance between data fusion scoring and Watvina scoring alone.

[0057] The results are as follows Figure 1 As shown in the figure, the performance of the fusion method (method 1) on the DUD-E dataset is significantly better than that of Watvina scoring alone. Specifically, on the top 1% EF, the average level of the data fusion method improved by X1, with a p-value below 0.001, indicating that the fusion method has statistical significance in screening ability. In addition, on the 5% and 10% EF, the average level of the fusion method also improved by X2 and X3, respectively, further verifying the effectiveness and superiority of the fusion method.

[0058] 3. Comparison of the screening performance of the data fusion method and Watvina scoring alone on the LIT-PCBA dataset

[0059] (1) Preparation of LIT-PCBA database: 15 protein targets, 7761 active molecules and 382674 inactive molecules were obtained from the LIT-PCBA official website (http: / / drugdesign.unistra.fr / LIT-PCBA). Proteins and small molecules were optimized and format converted using the rdkit2pdbqt.py script.

[0060] (2) Molecular docking using Watvina: Use Watvina to perform molecular docking on the optimized molecules to obtain the docked conformation and docking score.

[0061] (3) Re-scoring using CNNscore: Use CNNscore to re-score the conformation after Watvina docking.

[0062] (4) Fusion of Watvina scoring and CNNscore: Use the above method 1 for data processing.

[0063] (5) Calculate the enrichment factor to evaluate the screening performance of the data fusion method: Calculate the enrichment factors of the top 0.5%, top 1%, and top 5% using only Watvina scoring and using data fusion scoring.

[0064] The results are shown in Table 2. In the top 0.5%, top 1% and top 5% of the enrichment factors, the data fusion method is better than using only Watvina scoring, and the data fusion method improves the screening performance.

[0065] Table 2 Comparison of the performance of data fusion technology and Watvina screening alone based on PCBA dataset test

[0066]

[0067] 4. Validation of the data fusion method for TYK2 target

[0068] (1) Obtain active compounds from the ChEMBL database: 50 active compounds of TYK2 JH2 were obtained from the ChEMBL website (https: / / www.ebi.ac.uk / chembl / ). After optimizing these compounds, the format was converted using the rdkit2pdbqt.py script.

[0069] (2) Generate decoy molecules using the DUD-E database: Submit active molecules on the DUD-E website (https: / / dude.docking.org / ), generate 50 decoy molecules for each active molecule, and generate a total of 2500 decoy molecules. After optimizing all decoy molecules, convert the format using the rdkit2pdbqt.py script.

[0070] (3) Use Watvina to perform molecular docking on active molecules and decoy molecules: Use Watvina to perform molecular docking on the optimized molecules to obtain the docked conformation and docking score.

[0071] (4) Re-scoring using CNNscore: Use CNNscore to re-score the conformation of Watvina after docking.

[0072] (5) Fusion of Watvina scoring and CNNscore scoring: Use the above method 1 for data processing.

[0073] (6) Comparison of the screening performance of the data fusion method with that of using only CNNscore or Watvina scoring: Calculate the BEDROC (α=321.9, 80.5, 20.0) and the top 1%, top 5% and top 10% enrichment factors of the method using only CNNscore or Watvina scoring and the data fusion method.

[0074] The results are as follows Figure 2 It shows that for the TYK2 target, the data fusion method has better screening performance than relying solely on CNNscore or Watvina scoring.

[0075] Example 2: Method for screening TYK2 inhibitors based on fusion of deep learning scoring and traditional molecular docking scoring

[0076] 1. Acquisition and processing of TYK2 JH2 protein structure and small molecule compound library

[0077] (1) Protein preparation: Obtain the TYK2 JH2 protein structure (PDB ID: 6NZP) from the RCSB Protein Data Bank (PDB). Delete redundant chains, remove solvent molecules, add hydrogen atoms and charges. Convert the pre-processed protein to a new format using the rdkit2pdbqt.py script.

[0078] (2) Acquisition and processing of small molecules: 12 million small molecules were obtained from the TopScience database, and these small molecules were converted into different formats using the rdkit2pdbqt.py script.

[0079] 2. Molecular docking using Watvina

[0080] According to the position information of the crystallographic ligand of PDBID 6NZP, it is expanded in three coordinate directions Obtain the three-dimensional information of the docking box. Use Watvina to perform molecular docking between the small molecule and the JH2 protein structure to obtain the docking conformation and docking score. Eliminate small molecules with scores higher than -7.

[0081] 3. Calculation of relative binding energy

[0082] (1) Assembling proteins and small molecules and performing solvation treatment

[0083] Use PDBFixer (https: / / github.com / openmm / pdbfixer) to preprocess the protein file obtained in step 1 (1), including repairing missing residues, replacing non-standard residues, adding missing atoms and hydrogen atoms, etc. Use the Modeller module of OpenMM (https: / / github.com / openmm / openmm) to add an explicit water model. Use openbable to convert the pdbqt file of the small molecule conformation obtained in step 2 into sdf format. Merge the processed protein and small molecule, and remove any small molecule with a distance less than water molecules, avoiding the existence of water molecules colliding with small molecules.

[0084] (2) Energy minimization

[0085] The composite system was energy minimized using OpenMM's LangevinIntegrator to reach a stable energy state.

[0086] (3) Energy calculation

[0087] The potential energy of the complex system was obtained through OpenMM, using the GBSA implicit solvation model. Systems containing only the protein and the ligand were constructed separately, and then the potential energy of each was calculated. Finally, the relative binding energy was obtained by subtracting the sum of the total energies of the protein and the ligand when they were independent from the total energy of the complex system.

[0088] 4. Use CNNscore for re-scoring and screening

[0089] The conformations obtained in step 3 that satisfy the relative binding energy of less than -60 Kcal / mol were re-scored using CNNscore, and small molecules with CNNscores lower than 0.95 were eliminated.

[0090] 5. Fusion scoring and ranking

[0091] The docking score in step 2 is multiplied by the CNNscore in step 4 to obtain the fusion score. The top 40 small molecules are sorted according to the fusion score, as shown in Table 3. The chemical structures of these small molecules are shown in Figure 3 shown.

[0092] Table 3 Screening data of the top 40 small molecules

[0093]

[0094]

[0095] 6. In vitro activity test

[0096] The top 40 small molecules obtained in step 5 were tested for in vitro activity. The TYK2 JH2 Pseudokinase Domain Inhibitor Screening Assay Kit was used for biological evaluation.

[0097] Experimental conditions: A 384-well plate was used, and the system set volume was 20 μL. First, 40 compounds were preliminarily screened, and the small molecule concentration was set to 50 μM. 100 nM TYK2 JH2 protein was incubated with the inhibitor at room temperature for 10 minutes, and then 30 nM JH2 Probe 1 was added to initiate the binding reaction. After incubation at room temperature for 1 hour, the plate was read using a SpectraMax i3 plate reader (Molecular Devices) at an emission wavelength of 485 nm and an excitation wavelength of 535 nm. The results are shown in Table 3. Among the 40 compounds, STL250599 and STK595738 had good TYK2 inhibitory activity.

[0098] Further evaluation: The two screened compounds STL250599 and STK595738 were further evaluated. The two compounds were dissolved in DMSO and prepared into 11 half-log dilution series, with the highest concentration of 100μM. The final compound concentration range was 10μM to 169.2pM, and the final DMSO concentration was 1%. The biological evaluation process was the same as the initial evaluation. By drawing the IC50 curves of STL250599 and STK595738, as shown in Figure 4 As shown, the half inhibition concentrations obtained were 13.756 μM and 9.985 μM, respectively.

[0099] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for a person skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to replace some of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions claimed to be protected by the present invention.

Claims

1. A method for screening TYK2 inhibitors based on fusion of deep learning scoring and traditional molecular docking scoring, characterized in that: It includes the following steps: S1: Obtain the TYK2 protein structure and small molecules in the compound library, and pre-process them; S2: performing molecular docking on the small molecule pretreated in step S1 to obtain the docking conformation and its corresponding Watvina docking score; S3: Calculating the relative binding free energy of the molecular docking conformations that meet the Watvina docking score requirements in step S2; S4: re-scoring the molecular docking conformations whose relative binding free energy meets the conditions in step S3 based on deep learning, and outputting the deep learning scores; S5: performing fusion scoring on the molecules that meet the conditions and are re-scored in step S4, and selecting candidate compounds according to the fusion score ranking; S6: Conducting an in vitro activity test on the candidate compounds in step S5 to screen out TYK2 inhibitors with inhibitory activity.

2. The method according to claim 1, characterized in that In step S3, the small molecules that meet the Watvina docking score requirement are small molecules with a score no higher than -7.

3. The method according to claim 1, characterized in that: In step S4, the small molecules whose relative binding free energy meets the condition are those whose relative binding energy is less than -60 Kcal / mol.

4. The method according to claim 1, characterized in that In step S5, the small molecules that meet the re-scoring condition are those with a re-scored CNN score of not less than 0.

95.

5. The method according to claim 1, characterized in that The specific steps of the relative binding free energy in step S3 are: S31: Use PDBFixer to process the protein structure file pre-processed in step S1, including repairing missing residues, replacing non-standard residues, adding missing atoms and hydrogen atoms; use the Modeller module of OpenMM to add an explicit water model; use openbable to convert the small molecule docking conformation file obtained in step S2 into sdf format; merge the processed protein structure and small molecule, and remove any residues with a distance less than water molecules, avoiding the presence of water molecules that collide with small molecules; S32: using LangevinIntegrator of OpenMM to perform energy minimization on the complex system of step S32 to achieve a stable energy state; S33: Obtain the potential energy of the complex system of step S32 through OpenMM; construct systems containing only the protein and the ligand respectively, and then calculate their respective potential energies; and obtain the relative binding energy by subtracting the sum of the total energies of the protein and the ligand when they are independent from the total energy of the complex system.

6. The method according to claim 1, characterized in that In the step S2, the molecular docking score is Watvina; in the step S4, the re-scoring based on deep learning is CNNscore.

7. The method according to claim 1, characterized in that In step S5, the fusion score is obtained by multiplying the Watvina docking score of step S2 and the deep learning score of step S4.

8. A TYK2 inhibitor, characterized in that The TYK2 inhibitor is obtained by screening according to any one of claims 1-7.

9. The TYK2 inhibitor according to claim 8, characterized in that The chemical structure of the TYK2 inhibitor is specifically:

10. Use of the TYK2 inhibitor according to claim 9 in the preparation of a medicament for treating a TYK2-mediated disease.

Images