Method for constructing cross-species damage time inference model and damage time point prediction method

Abstract

The invention belongs to the field of forensic injury pathology, and particularly relates to construction of a cross-species injury time inference model and an injury time point prediction method. The sequential cross-species conservative molecules are applied to cross-species injury time inference based on the transcriptome, so that the accuracy and reliability of skeletal muscle injury recovery process inference are improved. According to the method, various analysis tools and algorithms are integrated, the cross-species conservative time sequence rule can be comprehensively mined, and a more comprehensive analysis view angle is provided. And the cross-species migration effect is reduced: through methods such as Combat, the cross-species migration effect of the biological data is effectively reduced, and the result accuracy is ensured. According to the method, a cross-species conservative time sequence rule is deeply excavated, the cross-species migration effect of biological data is reduced, the conservative property of biomarkers is deeply explored, and a comprehensive time sequence type cross-species research framework is established.

Description

Technical Field

[0001] The present invention belongs to the field of forensic injury pathology, and in particular relates to constructing a cross-species injury time inference model and an injury time point prediction method. Background Art

[0002] With the increase in traffic accidents, violent conflicts, accidental injuries and sports, skeletal muscle injuries are becoming increasingly common. Severe injuries not only significantly affect the quality of life of injured individuals, but also prolong recovery time. Therefore, accurate inference of injury time is not only crucial for the treatment of patients, but also has a profound impact on forensic medicine, sports medicine and public safety. However, due to ethical restrictions, we cannot directly study the repair process of human skeletal muscle injuries. Traditional research methods have low throughput and focus on a single model organism, making it difficult to reveal complex gene regulation mechanisms. Most studies on skeletal muscle injuries can only be conducted on animal models and are difficult to apply directly to humans. Therefore, how to migrate a large amount of experimental data from animal models to humans and achieve accurate prediction of human biological information is a pain point in current forensic research.

[0003] Most existing cross-species studies are limited to the comparison of a single species, often focusing only on biomarkers at a single time point, and lacking in-depth analysis of co-expressed genes, which limits the comprehensive understanding of the injury recovery process and makes it difficult to effectively transform data between different species across species. In addition, existing cross-species comparison methods often ignore the biological differences between species, resulting in insufficient comparability of research results.

[0004] Therefore, it is particularly important to break down the biological barriers between species and fill this gap through systematic cross-species research. Summary of the invention

[0005] In order to solve the current problem that the time of human skeletal muscle injury cannot be accurately predicted, the present invention provides a method for constructing a cross-species injury time inference model and an injury time point prediction method. Combining the advantages of transcriptomics, the present invention starts from cross-species homologous genes, deeply explores temporal cross-species conservative biological indicators, and uses methods such as ComBat to reduce the cross-species offset effect of biological data, and establishes a comprehensive temporal cross-species inference model to improve the understanding and prediction of the skeletal muscle injury recovery process. This innovative idea will not only provide scientific and objective methodological support for cross-species injury time inference, but will also promote the development of biomedicine and forensic medicine, and make up for the shortcomings of current research. With the increasing urgency of the need for injury time inference, the present invention will provide important technical support and theoretical basis for related fields.

[0006] The present invention is achieved through the following technical solution: a method for establishing a cross-species injury time inference model comprises the following steps: ① First, construct animal models of skeletal muscle contusion in at least two different animals at different injury times and collect tissue samples; extract total RNA from these tissues, sequence the samples, and finally generate transcriptome expression profiles at different animal skeletal muscle injury time points; ② Analyze and screen the orthologous genes or paralogous genes of different animals and humans to obtain homologous genes that are conservatively expressed across species; ③ The transcriptome expression profiles of different animal skeletal muscle injury time points were analyzed by weighted gene co-expression network analysis within each species; the intersection genes between species at each time point were merged to obtain the co-expressed gene set at cross-species injury time points; ④ Using the co-expressed gene set at the time of injury across species, we can obtain the protein interaction network of different animals and humans, perform deep mosaic of the cross-species protein expression network, and further obtain conservative biological indicators related to cross-species skeletal muscle injury repair; ⑤Introducing the public database human skeletal muscle injury dataset as an external validation of the cross-species model for skeletal muscle injury time inference; ⑤ Using the expression spectrum of the cross-species conservative biological indicators related to skeletal muscle injury and repair constructed above, a machine learning model for temporal cross-species conservative biological indicators was constructed.

[0007] As a further improvement of the technical solution of the method for establishing the present invention, before step ⑤, the cross-species conservative biological indicators related to skeletal muscle injury repair are merged to reduce the cross-species deviation of biological data.

[0008] As a further improvement of the technical solution of the method for establishing the present invention, in step ④, the Gosline network alignment tool is used to obtain the protein interaction network of different animals and humans.

[0009] As a further improvement of the technical solution of the method of the present invention, when reducing the cross-species bias of biological data, the Combat, Limma, Batch Mean centering or Genenorm method is used.

[0010] As a further improvement of the technical solution of the method of establishing the present invention, in step ②, when analyzing and screening species of different animals and humans, a pairwise gene comparison screening method is adopted to obtain an intersection and obtain homologous genes that are conservatively expressed across species. When multiple species are subjected to pairwise orthologous gene comparison screening, such as mouse-pig orthologous gene screening, mouse-human orthologous gene screening, and pig-human orthologous gene screening, and an intersection is taken to obtain the common orthologous genes of the multiple species under study.

[0011] As a further improvement of the technical solution of the method for establishing the present invention, in step ③, when performing weighted gene co-expression network analysis, the module with the strongest positive correlation with each damage time point is selected within each species, and then the intersection of the module gene sets at corresponding time points across species is taken to obtain the co-expressed gene set at the damage time point across species.

[0012] As a further improvement of the technical solution of the method for establishing the present invention, the machine learning model includes random forest, support vector machine, multilayer perceptron, and domain adaptive deep learning algorithm.

[0013] As a further improvement of the technical solution of the method of establishing the present invention, the order of steps ②, ③ and ④ can be replaced by ④, ②, ③, or ③, ④, ②.

[0014] It should be noted that when the sequence is ④, ②, and ③, when obtaining the protein interaction network of different animals and humans in step ④, the gene of the animal protein targeted is the transcriptome expression spectrum of different animals at the time points of skeletal muscle injury in step ①, and another set of cross-species skeletal muscle injury repair-related conservative biological indicators is obtained; then steps ② and ③ are performed.

[0015] As a further improvement of the technical solution of the method for establishing the present invention, in step ⑤, a human skeletal muscle injury dataset from a public database is introduced as an external validation of the cross-species model for skeletal muscle injury time inference.

[0016] The present invention also provides a cross-species injury time prediction method, which uses a machine learning model obtained by the above-mentioned method for establishing a cross-species injury time inference model, and the specific steps include: The cross-species conservative biological indicators related to skeletal muscle injury and repair of human skeletal muscle injury samples were input into the machine learning model to obtain cross-species predictions of the time point of human skeletal muscle injury.

[0017] Compared with the prior art, the cross-species injury time inference model and injury time point prediction method proposed in the present invention have the following advantages: 1) Improved scientificity and objectivity: Through advanced bioinformatics technology, starting from orthologous genes or paralogous genes, temporal cross-species conserved molecules are applied to cross-species injury time inference based on transcriptome, which improves the accuracy and reliability of the inference of skeletal muscle injury recovery process.

[0018] 2) Integration of multiple analysis tools: The method described in the present invention integrates multiple analysis tools and algorithms, which can fully explore the conservative temporal laws across species and provide a more comprehensive analysis perspective. Reducing cross-species bias effects: Through methods such as Combat, the study effectively reduced the bias effects of biological data in cross-species analysis and ensured the accuracy of the results.

[0019] 3) Establish a comprehensive research framework: By integrating a variety of analysis tools and algorithms through the method described in the present invention, starting from orthologous genes or paralogous genes, we can deeply explore the conservative temporal laws across species, reduce the cross-species bias effect of biological data, deeply explore the conservation of biomarkers, and establish a comprehensive temporal cross-species research framework. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 The implementation process of the method for establishing a cross-species injury time inference model involved in the present invention is demonstrated.

[0023] Figure 2 The correlation between cross-species co-expression modules and sample features is shown. The darker the color of the module, the stronger the correlation. In each unit, the top number represents the correlation coefficient (R value), while the number below represents the significance of the correlation (P value).

[0024] Figure 3 Deep chimera of the rat-pig-human cross-species protein expression network for the Gosline network alignment tool. Connecting lines represent similar expression between cross-species proteins.

[0025] Figure 4 This is the principal component analysis (PCA) diagram of the sample distribution of three species: rats, pigs, and humans before and after Combat treatment.

[0026] Figure 5 The external validation results of the cross-species model are shown in Figure 2. The confusion matrix results show the accuracy of injury time judgment, while the ROC curve evaluates the performance of the model. DETAILED DESCRIPTION

[0027] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the scheme of the present invention will be further described below. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0028] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present invention, rather than all of the embodiments.

[0029] The specific embodiments of the present invention are described in detail below.

[0030] 1. Experimental Animal Grouping This study selected 28 male Sprague Dawley (SD) rats aged 6 to 7 weeks (average weight 240 g ± 20 g), all provided by the Experimental Animal Center of Shanxi Medical University. These rats were randomly divided into a control group and an injury group, where the control group consisted of uninjured skeletal muscle rats, and the injury group included groups at 4 hours, 24 hours, and 48 hours after injury, with 7 rats in each group.

[0031] The present invention selected 4 9-month-old Bama pigs (weight 20.06±1.14kg), all provided by Beijing Shichuangshiji Miniature Pig Breeding Base. The Bama pigs were randomly divided into a control group and an injury group. The control group was a Bama pig with uninjured skeletal muscle, and the injury group included 4h, 24h and 48h groups after injury, with 1 pig in each group.

[0032] 2. Preparation of skeletal muscle injury animal model After fasting for 12 hours, the experimental rats were anesthetized by intraperitoneal injection of 3% sodium pentobarbital (0.13 ml / 100 g). A 500 g weight was dropped from a height of 50 cm to hit the right hind limb of the rat to establish a rat skeletal muscle contusion model (injury group). Then, sufficient food and water were given to continue feeding. At 4h, 24h and 48h after contusion, the experimental animals were euthanized with a lethal dose of 10% chloral hydrate (4 mL / kg body weight). The control group was not treated in any way and was handled in the same way. Subsequently, about 100 mg of muscle tissue samples were cut from the wound site of each rat. After removing the fascia, the samples were immediately frozen in liquid nitrogen and stored at -80 °C.

[0033] After fasting for 12 hours, the Bama pigs were fixed and atropine (1ml, 0.5mg) was injected intramuscularly at 0.1 mg / kg to observe the animal's reaction. When the animal's muscles were in a relaxed state, Shutai 50 (a mixture of pentobarbital and chlorpromazine) (50 mg / ml freeze-dried powder) was injected intramuscularly at a dose of 5 mg / kg. The animal's breathing, physiological state and body temperature changes were closely observed, and attention was paid to keeping warm. After the animal sedation took effect, the animal's front and rear limbs were fixed with ropes, the animal's mouth was fixed with a restraint belt, and the animal's abdomen was raised with a cushioning pad and a multi-layer foam plastic pad to fully avoid injury to the abdomen and internal organs during impact. A 2kg gravity hammer was dropped from a height of 1.66m to injure the pig's back skeletal muscles. Each pig was injured 4 times on both sides of the spine to make a pig skeletal muscle injury model, and then the pig was killed, and then the corresponding parts of the injury group were directly sampled. The control group animals did not have skeletal muscles bruised, and the same treatment was performed after being killed. About 100 mg of muscle tissue samples were cut from the wound of each Bama pig. After removal of the fascia, the samples were immediately frozen in liquid nitrogen and stored at −80 °C.

[0034] 3. Total RNA extraction and quality control of samples According to the manufacturer's instructions, approximately 50 mg of muscle tissue samples were placed in an RNase-free mortar and placed in a liquid nitrogen environment to keep the temperature low for grinding, and then lysed using RNAiso Plus9108 (Takara Bio, Shiga, Japan). The concentration (ng / ml) and purity of the extracted total RNA were determined by a microplate reader (Infinite M200 Pro; TECAN, Zurich, Switzerland), and the RNA quality was initially checked using a Nanodrop2000 spectrophotometer (Thermo, USA). Samples with an absorbance OD260 / 280 of RNA between 1.8 and 2.2 were used for subsequent experiments. The integrity of total RNA was identified by the Agilent2100Bioanalyzer using the Agilent2100Nano6000 analysis kit (Agilent Technologies, Palo Alto, CA, USA), and the RNA integrity number (RIN) of all samples was greater than 7.0, indicating that its quality was sufficient for poly-A selection and cDNA library preparation.

[0035] 4. Transcriptome library construction, second-generation sequencing and preliminary data analysis The RNA samples that had passed the quality control were sent to the work platform for the construction of the transcriptome library, and the NEBNextUltraTM RNA Library Preparation Kit (polyA Select, NEB, Beijing, China) was used for library construction. The library was preliminarily quantified using Qubit3.0 Fluorometer (Life Technologies). The insert size of the transcriptome library was measured by the Agilent2100 RNA6000 analysis kit, and the constructed library was accurately quantified using the Bio-RAD KitIQ SYBR GRN (Bio-Rad Laboratories Inc., Hercules, CA, USA).

[0036] According to the preparation process of mRNA high-throughput sequencing library, RNA is first reverse transcribed to synthesize the first and second chains, followed by the addition of adapters at the 3' end and the completion of the 5' end. After the completion of bridge PCR amplification, further purification, screening and sequence quality assessment are required to improve the accuracy of the library.

[0037] We performed paired-end sequencing on the HiSeq-2000 platform (Illumina Inc., San Diego, CA, USA) using a 150 bp length, a target coverage of 40-60M, and a sequencing length of PE150. The library of the muscle tissue sample was sequenced using the HiSeq PE Cluster Kitv4-cBot hS (Illumina, USA). After sequencing, bcl2fastq2 software was used for base calling, and the offline data was converted into raw sequence reads, and the results were saved in the FASTQ file format.

[0038] Next, a series of bioinformatics analyses were performed on the open source platform Chipster (version 3.16). The quality of the sequencing data was comprehensively evaluated using FastQC software, and HISAT2 (version 2.1.0) was selected to map the processed reads to the genome. Annotations were performed based on the Rattus_norvegicus.Rnor_6.0.954 and Sus_scrofa.Sscrofa11.1.95 files. HTSeq (version 0.6.1) was used to quantify and match the reads, using default settings and unique matching for the reverse strand to obtain gene-level counts and the original gene expression matrix for each time point after injury and the normal group. Finally, the original expression matrix was processed by the R Bioconductor package DESeq2 and DESeq normalization was performed. Genes with a total count of less than 100 in all samples were considered to be low-expressed genes and were removed.

[0039] 5. Bioinformatics Analysis (1) Mouse-pig-human homologous gene comparison After the above processing, the original expression profile was obtained, and then the mouse-pig-human homologous gene alignment was performed. The g:Orth tool was used to retrieve information from the Ensembl database to provide homologous genes, and then this information was used to identify homologous genes between rats, pigs and humans in all three species. A total of 6,367 orthologous genes in the three species were obtained. Subsequently, a weighted gene co-expression network analysis was performed.

[0040] (2) Weighted gene co-expression network analysis (WGCNA) We performed weighted gene co-expression network analysis (WGCNA) on 6367 homologous genes in rats and pigs. Using the WGCNA package in R Studio, we identified gene co-expression modules at each time point after injury. A soft threshold of β=22 was used in rats to obtain a scale-free topological structure, while a soft threshold of β=12 was used in pigs. The topological overlap was calculated using the adjacency matrix to measure the interconnectivity of the network. Using the dynamic tree cutting algorithm of WGCNA, 7 modules were identified in rats, including 6 meaningful modules and an undefined gray module; 16 modules were identified in pigs, including 15 meaningful modules and an undefined gray module. Subsequently, the R value and p value of each module were calculated using Spearman correlation analysis, and the module with the strongest positive correlation (i.e., the largest R value) at each time point was selected, which was called the High Correlated module (HCR) at each time point. Based on the results of the WGCNA analysis, we finally selected 11 co-expression modules, which were Figure 2 Indicated by different colors.

[0041] The genes of the corresponding modules at the same injury time point in rats and pigs were intersected, and then all the intersection genes were merged. Among them, the normal group (C) after skeletal muscle injury in rats and pigs had 19 cross-species conserved genes, the 4h group had 140 genes, the 24h group had 104 genes, and the 48h group had 22 genes. A total of 285 cross-species conserved co-expressed gene indicator sets were obtained.

[0042] 6. Further extraction of cross-species conservative biomarkers related to skeletal muscle injury and repair In the above study, we obtained 285 genes that were co-expressed and time-dependently expressed in the early stage of skeletal muscle injury in rats and pigs, and these genes were conserved in different related species. In addition, in order to study the interspecies conservation of these genes at the protein expression level, we obtained the protein interaction network of three species, rats, Bama pigs and humans, and used the PPI network local alignment analysis (GASOLINE) algorithm to achieve deep chimerism of cross-species protein expression networks, and explored possible conserved expression proteins in rats, pigs and humans. According to the default parameters, such as Figure 2 , the analysis revealed that there are 20 conservative protein interaction networks after skeletal muscle injury in three species: rat, pig and human ( Figure 3 Only the first two conserved networks among the 20 networks are displayed). After protein network alignment, the conservation of cross-species biological indicators is further strengthened. At this time, 105 conservative biological indicators related to skeletal muscle injury and repair in rats, pigs and humans are obtained.

[0043] 7. Incorporation of human skeletal muscle injury transcriptome data We found the human skeletal muscle injury datasets GSE45426 and GSE107934 from the GEO database, and selected 49 samples from the normal control group, 4 h post-injury group, and 24 h post-injury group as external validation of the cross-species model for skeletal muscle injury time inference.

[0044] 8. Reduce interspecies effects Due to the differences in gene expression levels between species, we used four methods (Combat, Limma, BatchMean centering, and Genenorm) to reduce species effects and allow the analysis of species data on the same dimension. The distribution of data at different time points after skeletal muscle injury in rats and pigs was shown using principal component analysis (PCA). Figure 3 The results of principal component analysis before and after the reduction of cross-species expression bias of biological data based on the Combat method are shown. After reducing the inter-species effect, samples at different time points after skeletal muscle injury showed temporal clustering, and the cross-species bias effect of transcriptome data was greatly reduced.

[0045] 9. Machine Learning Models Using the 105 cross-species conservative biological indicators related to skeletal muscle injury and repair that we obtained, we can achieve cross-species inference of the time point of skeletal muscle injury.

[0046] (1) Machine Learning Model Using the above-mentioned expression profiles of cross-species skeletal muscle injury and repair-related conservative biomarkers, we established a random forest (RF) cross-species machine learning model on the Anaconda (version 3) platform of Python. During the internal validation process, a five-fold cross-validation was used to evaluate the accuracy (Acc), precision (Pre), recall (Rec), and F1 score of the model. The sample data was randomly divided into a 70% training set and a 30% validation set, and the optimal parameters were determined by a grid search method (i.e., a strategy for finding the optimal parameters in each machine learning algorithm). In order to further test the reliability of our cross-species research strategy, skeletal muscle injury samples from rats and pigs were used for modeling, and human samples were used for external validation to achieve cross-species injury time point prediction.

[0047] Without combat to reduce the inter-species effect, the model of 105 temporal cross-species conservative biological indicators combined with random forest had an accuracy of 0.917 and an AUC value of 0.991 in internal validation. In external validation, the accuracy was 0.551 and the AUC value was 0.494. The confusion matrix showed that the external validation 4h samples and 24h samples were both misclassified as normal group C ( Figure 5 ) It is still not feasible to make cross-species inferences solely based on cross-species conserved biological indicators, and it is imperative to reduce the cross-species expression bias of biological data.

[0048] Next, after combat to reduce the inter-species effects, we found that the accuracy and AUC values ​​of 105 cross-species conservative biological indicators related to skeletal muscle injury and repair combined with the random forest model were both 1.000 in internal validation, and the accuracy and AUC values ​​were also 1.000 in external validation.

[0049] according to Figure 4 From the multi-species PCA graph, we can see that after Combat reduces the inter-species effect, the biological data is transformed from species clustering to time point clustering, and the inter-species effect is greatly reduced, which leads to a 100% accuracy rate in cross-species machine learning to determine the time point. This further shows that through our cross-species information transfer strategy framework, cross-species prediction of injury time points of multiple species can be achieved.

[0050] The above is only a specific implementation of the present invention, which enables those skilled in the art to understand or implement the present invention. Although detailed descriptions are given with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments, and they should all be covered by the protection scope of the claims.

Claims

1. A method for establishing a cross-species injury time inference model, characterized in that: The steps include: ① First, construct animal models of skeletal muscle contusion in at least two different animals at different injury times and collect tissue samples; extract total RNA from these tissues, sequence the samples, and finally generate transcriptome expression profiles at different animal skeletal muscle injury time points; ② Analyze and screen the orthologous genes or paralogous genes of different animals and humans to obtain homologous genes that are conservatively expressed across species; ③ The transcriptome expression profiles of different animal skeletal muscle injury time points were analyzed by weighted gene co-expression network analysis within each species; the intersection genes between species at each time point were merged to obtain the co-expressed gene set at cross-species injury time points; ④ Using the co-expressed gene set at the time of injury across species, we can obtain the protein interaction network of different animals and humans, and perform deep mosaicism of the cross-species protein expression network to further obtain cross-species conservative biological indicators related to skeletal muscle injury and repair; ⑤ Using the expression spectrum of the cross-species conservative biological indicators related to skeletal muscle injury and repair constructed above, a machine learning model for temporal cross-species conservative biological indicators was constructed.

2. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: Before step ⑤, cross-species conservative biological indicators related to skeletal muscle injury repair are merged to reduce cross-species bias of biological data.

3. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: In step ④, the protein interaction networks of different animals and humans are obtained from the String database, and the Gosline network alignment tool is used to achieve deep mosaicism of cross-species protein expression networks.

4. The method for establishing a cross-species injury time inference model according to claim 2, characterized in that: When reducing cross-species bias in biological data, the Combat, Limma, Batch Mean or Genenorm methods are used.

5. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: In step ②, when analyzing and screening species of different animals and humans, the g:Profiler R package is used to screen homologous genes between species, and finally a set of homologous genes with cross-species conservative expression of multiple species of animals and humans is obtained.

6. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: In step ③, when performing weighted gene co-expression network analysis, the module with the strongest positive correlation with each injury time point is selected within each species, and then the intersection of the module gene sets at the corresponding time points across species is taken to obtain the co-expressed gene set at the injury time points across species.

7. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: The machine learning models include random forest, support vector machine, multi-layer perceptron, and domain adaptive deep learning algorithm.

8. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: The order of steps ②, ③ and ④ can be replaced by ④, ②, ③, or ③, ④, ②.

9. The method for establishing a cross-species injury time inference model according to claim 1, characterized in that: In step ⑤, a public database of human skeletal muscle injury datasets was introduced as an external validation of the cross-species model for skeletal muscle injury time inference.

10. A cross-species injury time point prediction method, characterized in that: The machine learning model obtained by the method for establishing a cross-species injury time inference model according to any one of claims 1 to 9 is used, and the specific steps include: The cross-species conservative biological indicators related to skeletal muscle injury and repair of human skeletal muscle injury samples were input into the machine learning model to obtain cross-species predictions of the time point of human skeletal muscle injury.

Images