67 results about "Protein Databases" patented technology

A method and system are disclosed for identifying and/or locating complex patterns in an amino acid sequence stored in a computer file or database. According to an aspect of the present invention, techniques are provided to facilitate queries of protein databases. For protein descriptions received in response to the queries, embodiments of the present invention may scan the received protein descriptions to identify and locate Replikin patterns. A Replikin pattern is defined to be a sequence of 7 to about 50 amino acids that include the following three (3) characteristics, each of which may be recognized by an embodiment of the present invention: (1) the sequence has at least one lysine residue located six to ten amino acid residues from a second lysine residue; (2) the sequence has at least one histidine residue; and (3) at least 6% of the amino acids in the sequence are lysine residues.

The invention discloses a protein second-level mass spectrum identification method based on peak intensity recognition capability. The method comprises the following steps: firstly, virtualizing enzymatically hydrolyzed protein database sequence, establishing a peptide fragment database and a peptide fragment database index for peptide fragments subjected to enzymatic hydrolysis according to the mass number of the peptide fragments; then, finding out candidate peptide fragments conforming to the requirement from the established peptide fragment database according to the mass number of parent ions without charges in a to-be-analyzed experiment spectrum; then removing an isotopic peak and selecting an effective peak from the to-be-analyzed experiment spectrum so as to generate a theory spectrum of the candidate peptide fragments conforming to the requirement, counting peak intensity information of different ions, calculating the peak intensity recognition capability of different types of ions at different intervals, marking each candidate peptide fragment based on the peak intensity recognition capability, and selecting the peptide fragment with the highest mark as the authentication result of the experiment spectrum; and finally, performing quality control on the authentication result. The number of valid mass spectra and the number of valid protein peptide fragments, which are authenticated by the method, are both higher than those obtained by an existing algorithm; peaks can be selected dynamically; the running speed is high.

The invention provides an index acceleration method in scale protein identification, which comprises the following steps of: setting quality intervals for peptide sequences; setting the size of counting windows, and setting the number of the counting windows and the range of each counting window by combining the quality intervals; performing simulated enzyme digestion on protein database, and calculating the quantity of the peptide sequences in each counting window according to the quality of the peptide sequences obtained through the simulated enzyme digestion; obtaining the quantity of the peptide sequences which can be processed once in the memory of a computer according to the capacity of the memory of the computer, and obtaining a quality range section of the peptide sequences which can be processed once in the memory of the computer by combining the quantity of the peptide sequences in each counting window; performing the simulated enzyme digestion on the protein database, saving the obtained peptide sequences in one quality range section in the memory of the computer, and finishing the operations of sequencing, redundancy removal, and dictionary and inverted list establishment on the saved peptide sequences in the memory of the computer; and establishing a dictionary and an inverted list for each quality range section.

Feature of a compound is predicted by using information on interactions between substances. A database of interactions between compounds and genes/proteins is constructed on the base of information collected from bibliographic databases, gene/protein databases, and disease databases, and an interaction network is prepared by mapping the collected information to thereby enable prediction of the features of a compound.

The invention discloses an antibacterial peptide prediction method and device based on protein pre-training representation learning; the method comprises the following steps: S1, employing a pre-training strategy to carry out the word segmentation and covering of a label-free protein sequence from a protein database, and obtaining a pre-training representation learning model, carrying out pre-training of two tasks of covering a language model and sentence continuity prediction, capturing expressions of a word level and a sentence level, and helping the model to learn general structural features of a protein sequence; S2, for the antibacterial peptide pre-recognition and prediction task, changing an output layer of a pre-training model, and performing fine adjustment on the model by using an antibacterial peptide data set with a label to generate an antibacterial peptide prediction model; and S3, according to the antibacterial peptide pre-identification and prediction task, adopting an antibacterial peptide prediction model for identification, and outputting a prediction result. Pre-training is applied to the field of antibacterial peptide recognition and prediction, and an efficient antibacterial peptide prediction model is established based on a known antibacterial peptide sequence with small data volume and unbalanced distribution.

The invention relates to a high-throughput retrieval method for drug targets, and belongs to the field of bioinformatics. The high-throughput retrieval method includes the steps that a drug and target complex serves as reference, a drug combining bag is defined, all fragments in the combining bag are represented with protein structural fingerprints, and the protein structural fingerprints include amino acid sequences, protein folding shape codes, physicochemical properties and vector coupling; the digital drug combining bag is input, a global known protein structure database is retrieved to perform fingerprint comparison and quantitative evaluation, and protein structures are arrayed in the sequence of fingerprint similarity from high to low; structural protein is selected as possible target spot regions, wherein similarity scores of the protein folding codes and similarity scores of the amino acid physicochemical properties reach top two thousand at the same time, and possible target protein of drugs is analyzed and predicted. The high-throughput retrieval method can be applied to secondary development and research of the drugs, and new effects of the approved clinic drugs are developed by finding the new targets.

The invention provides a protein drug binding site prediction method based on deep learning, and the method comprises the following steps: 1, selecting a plurality of proteins in a protein database toform a training set, a plurality of proteins to form a verification set, and a plurality of proteins to form a test set, wherein the training set being used for training a training model; step 2, carrying out feature extraction and label extraction on the protein database through the trained training model to obtain data, finishing training of a neural network and obtaining a prediction model; and 3, inputting the new protein into the prediction model, and positioning and predicting the position of the binding site. The method has the advantages that the forming factors of the binding site are comprehensively considered, and the binding site is positioned and predicted based on deep learning.

The invention relates to a peptide identification method based on subset error rate estimation. The peptide identification method comprises the following steps: 1, analyzing a peptide sample to be identified by a mass spectrometer to generate a tandem mass spectrum; 2, searching a target-bait protein database containing a target peptide sequence in the tandem mass spectrum, and sorting obtained peptide identification results according to scores from high to low; 3, setting a score threshold value x, and estimating the error rate FDRk(x) of a type k peptide identification subset, the score of which is higher than x, by a transferring FDR (False Discovery Rate) method; 4, finding the minimum value of x by adjusting the score threshold value x to enable the estimated FDRk(x) to be less than a given error rate control level alpha, so that the obtained type k peptide identification result with the score higher than x serves as an acceptable reliable identification result. The peptide identification method provided by the invention estimates the subset error rate through the transferring FDR method and obtains the reliable peptide identification result through the subset error rate, thus having high identification accuracy.

The invention relates to a protein structure prediction method and device based on a multi-task time domain convolutional neural network. The method comprises the steps of: obtaining a target gene sequence and a protein database; establishing a DNA RNAamino acid ternary sequence data set corresponding to each protein according to the genetic code table and a protein database; establishing a multiple regression equation according to the residue depth and physicochemical properties of amino acids in the protein database to obtain statistical depth characteristics of each protein; clustering theternary sequence data set and mapping the ternary sequence data set into a multi-dimensional feature vector; taking the multi-dimensional feature vector and the statistical depth feature of the protein as the input of a multi-task time domain convolutional neural network, and training the multi-task time domain convolutional neural network; and predicting the protein structure by utilizing the statistical depth characteristics of the protein. According to the invention, the statistical depth characteristics of the protein are combined with the multi-task time domain convolutional neural network, so that the complexity of the model is reduced, and the generalization and the fitting degree are improved.

The invention discloses a protein structure prediction method. The method comprises the following steps: extracting a target sequence from a to-be-detected protein file; matching the target sequence in a protein database with a known structure to find a matching sequence; obtaining a matching structure of the matching sequence according to the matching sequence; constructing an initial three-dimensional structure model of the target sequence based on the matching sequence and the matching structure thereof; combining the unmatched sequence segment of the target sequence with one part of the adjacent matched sequence segment to form a sub-target sequence; searching a matching subsequence and a structure of the matching subsequence of the sub-target sequence in a protein database with a known structure; and filling the missing part in the initial three-dimensional structure model according to the searched matching subsequence and the structure thereof to obtain the three-dimensional structure of the to-be-detected protein file. By adopting the protein structure prediction method provided by the invention, the structure of the protein can be more accurately predicted.

The invention provides a method for obtaining a tumor urine protein marker and an obtained stray urine protein library related to a tumor. The method comprises the steps that based on a built quantitative reference range of human urine protein in a healthy human urine protein database, a mode of hypergeometric distribution detection is used for screening stray protein as the tumor urine protein marker from a urine proteome dataset of a tumor patient, and the stray urine protein library related to the tumor is established. The method for obtaining the tumor urine protein marker and the obtainedstray urine protein library related to the tumor can better eliminate the interference of physiological fluctuation and inter-individual differential protein in the process of finding a urine proteinbiomarker.

The invention discloses a compression and clustering-based batch protein homology search method and belongs to the cross field of computer application technologies and bio-technologies. The method comprises the steps of firstly performing compression operation on a query sequence and a protein database through redundancy analysis and redundancy removal processes by fully utilizing sequence similar information existent in a protein database sequence and the query sequence; secondly performing similar sub-sequence clustering on the compressed protein database; thirdly performing a search by utilizing a mapping principle based on the clustered database to discover potential results, and establishing an executable database according to the found potential result set; and finally performing a homology search in the executable database to obtain a final homology sequence. According to the method, the homology search is performed in the established executable database, so that the time for repeated sequence comparison and gapless expansion is greatly shortened.

The invention provides a protein structure prediction method, a protein structure prediction device and a medium. The protein structure prediction method is applied to the computer equipment, the computer equipment comprises a CPU and at least one GPU, and the method comprises the following steps: obtaining a target protein sequence of a to-be-predicted protein structure. And in the CPU, according to the sequence length of the target protein sequence, determining an alignment quantity threshold value of a matching sequence corresponding to the target protein sequence. And comparing the target protein sequence with a plurality of protein sequences in a preset protein sequence library according to the comparison quantity threshold, and determining a matching sequence corresponding to the target protein sequence. And determining a matching structure corresponding to the matching sequence in a preset protein structure database. And inputting the matching sequence and the matching structure into a protein structure prediction model preset in a GPU for protein structure prediction, and obtaining a protein prediction structure corresponding to the target protein sequence. The memory occupation of the GPU can be reduced, the operation speed of the GPU is improved, and the prediction rate is accelerated.

The invention relates to a method for carrying out large-scale proteomics identification based on a silkworm tissue sample. The method comprises the following steps: pre-treating a domestic silkworm proteomics sample, carrying out graded peptide fragment mass spectrum online detection and constructing a silkworm protein database. By optimizing a protein extraction method, a peptide fragment is divided into 8 grades by adopting a high pH (Potential of Hydrogen) grading method and silkworm fat body samples can be extracted as many as possible, so that the protein identification quantity is improved; a condition that a sample spraying needle is blocked, caused by the fact that a feeding amount is too great, is prevented through optimizing a sample feeding amount and chromatography gradient time; the detection time is shortened through optimizing the chromatography gradient time; a Streamline database containing 21,878 protein sequences is established; the database can be used for identifying more protein quantity and redundant sequences are removed by the database; later-period proteomics data analysis is facilitated. According to the method provided by the invention, a stable and efficient domestic silkworm proteomics identification platform is established and the method has important meaning on silkworm proteomics large-scale identification.

The invention discloses salinimonas profundi 13199. The salinimonas profundi has the preservation number being CGMCC1.17396 and a classification name being Salinimonas profundi. The strain is a new bacterial species derived from deep sea, and contains a special CRISPR-Cas system and a large number of heavy metal resistance genes. The sequence homologies of the Cas protein and a protein which is included in a high-quality protein database UniProtKB/Switch-Prot subjected to a certain degree of functional researches are lower than 65%, and possibly have different activities and characteristics. The strain Salinimonas prodi 13199 and the CRISPR-Cas system thereof are novel materials for developing and utilizing the CRISPR-Cas system and performing related researches, and furthermore, the strain has a certain application prospect in the aspects of heavy metal polluted environmental modification preparations.

The invention discloses a protein folding identification method based on triple loss, which comprises the following steps of encoding protein by using one-hot encoding, inputting the encoded protein into an SSA program to obtain a contact graph between protein residues, and using the contact graph as input data. inputting the input data into a pre-trained deep learning framework, wherein the output of the network is the characteristic that the protein is specific to folding identification, comparing characteristics of the query protein with template proteins of known protein folding categoriesin a protein database, and assigning the folding category of the template protein closest to the query protein to the query protein. According to the method, the training thought of triple loss is used for reference, so that protein structures of the same class are closer, protein structures of different classes are farther, feature expression of protein has higher discriminability, and the recognition efficiency is higher.

The invention obtains the specific expression protein of wheat salt tolerant mutant, and finds by mass spectrum identification, Web protein database search and bio-information technique that the protein has high homology with one assumed protein on the 4th rice chromosome. Accordingly, it designs primer, amplifies by PCR to obtain the corresponding wheat salt tolerance gene with 153 amino acids and 462bp length; clones the gene to construct pronucleus expression vector to express out a specific protein with molecular weight as 16.8KD and improve the salt tolerance of escherichia coli obviously. This invention settles foundation for wheat salt tolerance mechanism.

The invention discloses a method for predicting antibacterial peptides of lactic acid bacteria based on a graph neural network. The method comprises the following steps: establishing a positive sample by searching known antibacterial peptides of lactic acid bacteria, establishing a negative sample by collecting sequences with the length of 5 to 255 from a protein database, and removing redundant sequences and similarities; performing feature extraction according to the positive and negative samples to obtain a feature vector and an initial input graph, and establishing a graph neural network model on the basis; through training, evaluation and loop optimization of the graph neural network model, determining parameters such as the optimal layer number, the optimal training round number and the learning rate of the graph neural network; and finally, predicting data of strains suspected to have antibacterial activity according to the graph neural network model. By adopting the method for predicting the antibacterial peptides of the lactic acid bacteria, wet experiment screening in a laboratory is replaced by computer model prediction, the judgment time of the protein sequence of the antibacterial peptides of the lactic acid bacteria is shortened, accurate and efficient batch identification is realized, and an effective alternative method is provided for screening lactic acid bacteria strains with antibacterial characteristics.

The invention discloses a protease substrate screening method based on solid-loaded mixed protein as a screening database. Specifically, mixed protein is bonded onto a solid phase carrier by chemical action to use as the screening database, the protein database is incubated with protease, peptide fragments which are produced by enzyme digestion of substrate protein of the protease in the protein database enter into a solution, non substrate protein cannot be enzyme digested, so that the non substrate protein is completely retained in the solid phase carrier, after solid-liquid separation, qualitative and quantitative identification of the substrate fragments in the solution is performed by bio mass spectrometry so as to obtain the specific protease substrate protein information. The protease substrate screening method is a method using highly specific effects of the protease and a substrate thereof and simplicity of the solid-liquid separation to screen the substrate protein from the mixed protein database.