Disease-related gene screening method and animal model construction method

Abstract

A disease-related gene screening method, comprising the following steps: step 1), constructing a data set of a gene expression profile comprising diseased and non-diseased samples; step 2), on the basis of the data set in step 1), establishing a linear model between each piece of gene expression data in the gene expression profile and whether the gene expression data is diseased; and step 3), on the basis of the slope of the linear model for each gene obtained in step 2), screening out differentially expressed genes, the differentially expressed genes being disease-related genes. Known regulatory network information is combined with full genome expression data by means of a network model of a genotype-phenotype relationship, and the causal relationship from gene to gene product to phenotype is quantified by means of a network mathematical model, so that the causal influence of monomorphic or polymorphic genes is simulated.

Description

A method for screening disease-related genes and a method for constructing an animal model

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application filed with the China Patent Office on October 16, 2023, with application number 202311341564.7 and invention name “A method for screening disease-related genes and a method for constructing an animal model”, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates to the field of biotechnology, and specifically to a method for screening disease-related genes and a method for constructing an animal model. Background Art

[0004] Preeclampsia is a common type of hypertensive disorder during pregnancy. Internationally, it is defined as new-onset hypertension after 20 weeks of gestation accompanied by proteinuria or maternal organ or uteroplacental dysfunction. Preeclampsia puts both mother and fetus at serious risk. For the mother, this leads to a 2-4 times increased risk of long-term hypertension, a 2-fold increased risk of cardiovascular death and major cardiovascular events, and a 1.5-fold increased risk of stroke. For the fetus, this leads to antenatal risks of intrauterine growth restriction (IUGR), oligohydramnios, preterm birth (most commonly due to human factors), placental abruption, fetal distress, and fetal death.

[0005] Preeclampsia progresses in two stages: Stage I: early placental dysfunction and Stage II: late multi-organ dysfunction. Major risk factors for preeclampsia include a history of preeclampsia, chronic hypertension, gestational diabetes, multiple pregnancies, antiphospholipid syndrome, and obesity. The severe clinical phenotype of preeclampsia (Stage II) includes gestational hypertension, proteinuria, renal complications, liver damage, neurological complications, hematologic complications, and uteroplacental complications.

[0006] Other factors that increase the risk of preeclampsia include systemic lupus erythematosus, history of stillbirth, fasting, previous placental abruption, assisted reproductive technology, chronic kidney disease, and advanced maternal age.

[0007] Table 1: Standard descriptions of pre-eclampsia.

[0008] Additional risk factors include polycystic ovary syndrome, sleep-disordered breathing, and various infections such as periodontal disease, urinary tract infections, and Helicobacter pylori. Genetic markers for preeclampsia have been explored. A genome-wide association study of 4,380 patients with the condition and 310,238 controls identified a locus near the FLT1 (FMS-like tyrosine kinase 1) gene as involved in the development of preeclampsia. Several studies have reported on the risk factors and pathogenesis of preeclampsia, such as those in the literature. The pathogenesis of preeclampsia has been elucidated. Increased levels of angiogenic markers, such as soluble FMS-like tyrosine kinase-1 (sFlt-1) and soluble enegrin (sEng), are also implicated. sFlt-1 reduces levels of vascular endothelial growth factor (VEGF) and placental growth factor, both of which are important regulators of endothelial cell function. sEng is a cell surface receptor that reduces levels of transforming growth factor (TGF)-β, which is responsible for endothelial cell migration and proliferation. All of these factors contribute to endothelial dysfunction, vasoconstrictive states, oxidative stress, and microembolism involving multiple organ systems. However, existing technologies have not focused on the genetic and molecular functions underlying severe preeclampsia. Next-generation sequencing, a massively parallel sequencing technology that enables affordable analysis of large genomic regions, is a promising tool for studying the genetic influence on preeclampsia. Utilizing genome-wide gene expression profiles associated with clinical phenotypes will demonstrate the unbiased nature of systematic analysis and identify key disease genes for severe preeclampsia, paving the way for further drug development.

[0009] Current approaches to managing preeclampsia include preconception counseling, perinatal blood pressure control and monitoring, perinatal aspirin therapy for high-risk women, betamethasone for women younger than 34 weeks, parenteral magnesium sulfate, and postpartum blood pressure monitoring. Prompt delivery of the fetus is currently the only established treatment. The International Society for the Study of Hypertension in Pregnancy recommends exercise to reduce / prevent the development of preeclampsia, but this is not consistently supported by multiple studies. For the treatment of severe preeclampsia, antihypertensive medications have shown promising results in lowering blood pressure. Anticonvulsants, such as magnesium sulfate, pravastatin, and metformin, have been proposed for preventing seizures in preeclampsia. Pravastatin has shown promising results in mouse models, and a human clinical trial is underway. Corticosteroids promote lung development before delivery. Preliminary data on the effects of metformin in obese pregnant women appear promising. Severe preeclampsia is a complex, multifactorial condition and a leading cause of mortality during pregnancy, with no effective treatment strategies currently available. This is largely due to the lack of reliable preclinical models that replicate the severe human disease of preeclampsia.

[0010] In summary, there is an urgent need for a method to identify the key disease genes of preeclampsia and further construct a model to replicate severe human preeclampsia.

[0011] Summary of the Invention

[0012] Therefore, the technical problem to be solved by this application is to provide a method for identifying key disease genes for preeclampsia. This application adopts the following technical solution:

[0013] A method for screening disease-related genes comprises the following steps:

[0014] Step 1) constructing a dataset containing gene expression profiles of diseased and non-diseased samples;

[0015] Step 2) establishing a linear model between each gene expression data in the gene expression profile and whether the disease occurs based on the data set in step 1);

[0016] Step 3) According to the slope of the linear model of each gene obtained in step 2), differentially expressed genes are screened out, and the differentially expressed genes are the disease-related genes.

[0017] Optionally, in step 1), multiple data sets are constructed; in step 2), a linear model is established for each gene expression data in each data set and whether it is a pre-eclampsia sample phenotype; in step 3), genes that are differentially expressed in each data set are screened according to the slope of the linear model of each gene.

[0018] Optionally, step 2) specifically includes the following steps: using each gene expression data in the gene expression profile as a predictor variable and whether the disease is present as a response variable; and setting the value of the diseased sample phenotype to 1 and the value of the non-disease phenotype to 2 to establish a linear model.

[0019] Optionally, step 3) specifically includes the following steps: expanding the slope value of the line model of each gene to the range of [-1, 1] to obtain a score value for each gene, comparing the score value with a threshold θ representing a significant difference, and genes with a score value greater than the threshold are differentially expressed genes.

[0020] Optionally, the disease is pre-eclampsia.

[0021] Optionally, the data set of step 1) further includes clinical phenotype data of pre-eclampsia samples and non-pre-eclampsia samples;

[0022] The clinical phenotype data includes at least one of diagnosis result, average uterine PI, average umbilical cord PI, maximum systolic blood pressure, maximum diastolic blood pressure, proteinuria pattern, HELLP diagnosis, and gestational age;

[0023] The step 2) further includes establishing a linear model between each gene expression data in the gene expression profile and each clinical phenotype data;

[0024] After screening the differentially expressed genes in step 3), a scoring matrix is ​​established; the rows and columns in the scoring matrix represent the differentially expressed genes and the types of clinical phenotypes, respectively, and the slopes of the linear model obtained in step 2) are filled in the corresponding positions in the scoring matrix; the scoring matrix is ​​subjected to a dichotomous cluster analysis to classify the clinical phenotypes.

[0025] Optionally, pathway analysis of the disease-related genes screened is also included.

[0026] The present application also proposes a gene set obtained based on the above screening method, wherein the gene set comprises the following genes: ANG, C7orf49, DONSON, SLC38A10, DSCR4, HERPUD1, PDLIM2, HSPA2, LACTB, ADAM17, ANO8, GNPTG, DDIT4L, TMCC3, PMEL, PGRMC1, VASN, PXN, PGK1, FGD6, BCAP31, KRT15, ​​SPOCK2, VWA1, MT1F, BIN2, DDR1, RPS19BP1, PJA1, DAB2IP, KIAA1217, RANBP2, CD9, ISM2, CCDC127, TMEM141, DDX58, PF KFB3, NEBL, HPCA, EPHB3, MED12L, SEMA3G, REEP4, HSPG2, CCDC115, GALNT18, LRRC32, GRHL3, KRT7, KLHDC7A, M ETTL6, FABP4, SH3BP2, TNNI2, LAIR2, ARSK, HRASLS2, LAMC2, FAM219A, GPER, GREB1L, CLDN5, MIDN, DNAJB2, I TGB6, MAPK8IP3, FPGS, SIPA1L1, VEGFC, MAP3K8, OXGR1, SPICE1, PTPN12, ADRA2C, RIOK1, DLX5, SHANK2, ATP6V 0E1, JUNB, NANS, LIMK1, GFOD2, ZFP36L1, CKMT2, TPM1, SFR1, STC2, PTGES, ATG16L2, FBLIM1, UBR1, SLC2A5, CD K19, PWWP2B, SSR1, PPARG, CEBPA, DEGS1, PFKL, TCF7L2, TNR, ERP44, GSE1, DCP2, ABCA12, GPR32, IRF9, TNIP2, ZBTB2, FOXJ3, CA4, JUP, CAV3, FGF22, PDIA5, KCTD11, FAM8A1, BCAR3, AP3S1, IFI44L, STK11IP, DVL1, LBH, SNX9, FN1, NLRP2, MTMR4, TMEM179B, VIT, ANKRD23, ANXA4, ELMO 3. PFKP, SSFA2, SLC11A2, MAP3K9, B9D2, POLR2H, INSIG1, FOSL2, DOCK5, UBASH3B, SLC35F2, SYN3, MAPK11, IL1R2, DGKZ, HMOX1, ELF3, SNX33, LY6D, AHNAK, GATA3, MAN1C1,<h2 style=";text-align:left;direction:ltr">ABCA7, CXXC1, KCTD9, AVL9, ISLR2, DNAH11, PREP, JPH1, ASTE1, IGSf3, CLDN9, NARF, SH3RF1, UPK1B, BMPR2, PDK1, BCAT2, SWAP70, SERPINE1, SLC2A1, EB I3,PMEPA1,TRIP10,SLC28A1,AMOTL2,MORF4L2,RHOB,SLC38A3,TIMP3,NLR P7,CITED2,C5orf30,PFKFB4,SLC2A3,BASP1,FAM184A,HPCAL1,RIMKLA,FBX O22,PLAC8,EPHA1,KLF4,ANKRD13B,CYP11A1,STX1A,MFAP5,GLA,ELF4,SUV 39H1,SLC1A6,EFHD1,HN1L,FAM156A,P2RY2,RAB3D,IRF2BPL,CSNK2A2,RHOF ,OAS3,TGM1,HK1,STK24,RASSF7,CCRN4L,KIF17,C14orf37,CCDC113,MYCN ,SLCO4A1,AQP1,RARRES3,LYN,MED13L,BTNL9,MBD2,ASCL2,ALPK3,KIT,PRR 15,FAM43A,AGRN,CRKL,PAM,RHOU,MIIP,PPP1R16B,FOXJ2,PTPRU,FKBP2,M MP14,PLEC,CRH,IL1RAP,TMEM59,PADI1,VTCN1,GLIS3,PLBD1,PNCK,BAD,R NF24,SUN2,SPTLC3,ING1,CPOX,ARNT2,GREM2,UGP2,DNAJC1,ACSF2,HEY1, USP18,EFNA1,IGSF8,FRMD4A,MAML2,SLC16A1,KIAA1919,HPS3,PPP1R3B,TN FAIP8,RAB25,MIF,FKBP8,TREM1,FAM120AOS,PPL,ITGA5,HMGN3,EPS8L1,R ASSF8,F11R,C4orf3,KIF2A,MXI1,MTUS1,GALE,GAPDH,RUSC2,HCAR2,TPI1, LTBP1,GPX3,KLF3,TES,JPH2,TMEM66,SLC2A14,CDO1,NEK6,PARD3,RASGRF 2,HMGB3,SEC23A,TMEM184A,GSN,RAB8B,ILVBL,SLC16A3,MOB3A,SERPINB9,<h2 style=";text-align:left;direction:ltr">EDEM2,SFRP1,PPP1R12C,TRIM24,ANKRD37,ARID3A,LEPROTL1,KDM3A,CSNK 1E,FHL2,CCSAP,NEURL,S100A6,CXCR6,FUT11,DSP,EAF1,GALK1,FAM120A, ERRFI1, VWCE, ALDOA, LRRC8A, AK4, ZMYND8, EZR, TACC1, NTRK2, ADAM9, ADORA2A, HTRA1, GUCA2A, GPR17, ARDC3, EGLN3, PRKAB2, SERPINA3, PRSS22, EGLN 1,C5orf46,MID1,GPT2,PTBP3,LRRC1,NRIP1,PAPPA2,CCBP2,ARHGEF4,MGA T3,AIF1L,SLCO2A1,KRT17,NOG,PLOD2,DUSP1,PKM,ASPHD2,PLIN2,CYP2J2 ,QPCT,HILPDA,RDH13,NEK11,CORO2A,GRHL1,SCARB1,GALNT2,DNM2,PIK3A P1,SYDE1,PRRG1,PIK3R1,LIMCH1,UBE2Q2,GDPD3,HK2,MLL5,TLE6,GPR146, BTD, PHYHIP, MYOF, PRKD3, ZCCHC2, CALM1, SIGLEC6, BCL6, LDHA, RASEF, HEXB, ENG, INHA, INHBA, MME, PVRL4, BACH1, PLA2G16, TSC22D1, TRIM14, TBC1D2 2A, CCBL1, WDR60, LIMD1, NPNT, TMEM45A, SEMA4C, IPMK, ERGIC1, MTSS1L, BHLHE40, SLC6A8, RFK, CST6, DIO2, C12orf75, PROCR, PIK3CB, ERO1L, LEP, SFXN 3,SH3BP5,MYO7B,MICAL3,KRT19,HTRA4,TPBG,SPAG4,EFNB1,SASH1,OCRL, SMARCA1,P4HA1,NDRG1,SH3PXD2A,COL17A1,FSTL3,FLNB,FLT1,ELMO1,TRPV 6,EPHB6,HSD17B1,ARL4C,TMEM139,PKD2,TRIM40,SEMA3D,PDE7B,SLC20A1 ,HIST1H1T,ANXA6,BTG2,FBN2,ASAH1,SLC25A27,RAB15,ZC2HC1A,TMEM136,<h2 style=";text-align:left;direction:ltr">KLRD1, TMEM168, CAP2, HSD11B2, SELE, STON2, SCN9A, ABCD3, PISD, PPAP2B, F13A1, CYTH3, OLFML3, FAM101B, SOD1, CSHL1, LRRC4, SNX27, CYBASC3, SOX5, ABAT, ADAMTSL1, C18orf8, LEPREL1, LRIG3, B3GNT2, ABCG2, GRIK2, DISP1, CNTN5, GABARAPL1, CATSPERB, ARPC5, PLCXD2, S LC16A10,DAB2,TKT,CD28,MORC4,EGFL6,CADM3,KCTD3,SLCO2B1,SY NPO2L,SLC23A2,PXK,SLC9A9,PDE5A,IPO13,GCLM,OPRK1,ATP8B4,DN MT3A,LAMA2,IL12RB2,GSTA3,C12orf39,RNF114,MTSS1,TMEM133,C DYL2,MOB3C,CLDN16,C10orf10,ECM1,ACSS1,NPC1,IGF1,SLAMF1,AT P1A1,HSD3B2,NUFIP2,LRRN3,ADAP2,MAF,LSS,EML2,POSTN,RAPH1, SPATA6,LAMC1,APLN,KL,ANK3,TRAK2,WWC1,CPEB4,MAPK8,GNG2,GOT 1,FAM20B,ZNF555,AKIP1,SPON1,TLR7,FAM84B,PRKCE,KCNIP3,RCA N1,HDAC9,RALYL,THSD4,ESRRG,HNRNPA0,IL36RN,CLIC5,VNN3,ST5, ACO1,LGI2,LCP1,CYBB,TXNRD1,KCNIP1,SH3BGRL2,TRAF3,ZNF554, MRS2,KLHL3,ADCY7,RAD54L2,HPGDS,PLEKHO2,UNC5C,HLF,RAP1GAP2 ,UBALD2,CD200R1,ATRNL1,ENPP1,LGALS14,NFIA,ZNF682,PSEN2,D OCK8,C5,SIK3,MKNK1,FES,BMX,SPRED1,ZBTB39,FOXRED2,ZMYND11, GPR34,TNFRSF11A,UBB,OR51B5,FOLR2,SLC19A3,CLCN5,TGFBR3,ABHD3,PMAIP1,TOM1L1,OR51B4,MRPS30,NFE2L2,<h2 style=";text-align:left;direction:ltr">SLC2A12,SCG2,CYSLTR1,MYO1A,VCP,OTUB2,NCEH1,SLIT2,KIAA0226L,PIT PNA,LYPD6,PLAC1,HMGN2,HIST2H2BE,SUPT7L,DCAF11,PCBD1,GRINA,CABL ES1, C5orf42, ALDH4A1, RAB5C, FRZB, SFXN4, CHKA, APMAP, SLC26A2, CERS2, ATXN2, MAP7D2, NEDD4L, ITPR1, ENPP2, DSN1, IQGAP2, FRY, LYVE1, MB21D2, LA MB2,CARD16,GSTK1,GSK3B,MATN2,MFF,NALCN,PIKFYVE,KIRREL,TBC1D9,P DPN,TMEM55A,ICA1,APBB2,ARHGAP17,PIK3CG,BMP5,SOX6,PSMC5,AFF3,FGF 10,GUCY1A3,RBM47,PRKG1,GLUL,ASB2,STRADA,CCND2,FILIP1,ST3GAL4,L ANCL2,SLC37A1,PDZRN3,TSPAN33,VAV1,DEPDC5,PER3,WNT2,CLDN1,RUNX1T 1,AMPD3,OR51B2,CYP4V2,IDO1,PC,XPO6,ZNF583,ZNF641,TYRO3,CYTH4,N SMAF,NCKAP1L,PTPN1,FRMD6,SRPX,ELK1,ACSS3,PIAS2,KLF12,TM9SF3,STX BP6,BHLHE41,FLRT3,PTPRD,TMEM106C,ZNF610,PHTF2,ZEB2,HSPB11,P2RY 1,RAB3IL1,CRYAB,ABCA11P,LAMC3,TMPRSS3,FMO1,SLC4A4,BMP4,RBBP4,DU SP4,CXorf57,HYOU1,KLHL42,ANGPTL2,PARP6,AGPAT5,WDSUB1,ABCA9,ATP 5B,C11orf63,DUS2L,PIGX,ZNF542,DACH1,LRRTM2,LRRC39,SEC22C,LINC00 472, TSPAN7, GABRA3, CHN1, ASPA, FXYD3, PRTFDC1, GUCY1A2, ANKS1B, SLC25A6, AIF1, PDHA1, BIN1, PCOLCE, DLG2, VTN, FXN, CTLA4, SLC17A5, APBA1, NINTC2N, PTP4A1, TXK, MAD2L2, FURIN, ERCC6, SGCB, OSBPL11, RDH10, LONRF2, TMEM176A, EID1, VAV3, SLC35F6, K CNJ5, CX3CR1, TMEM86A, NPL, MAT2A, HDAC3, GRHPR, SPEF2, ST6GAL1, BNC2, PRKCD, CLTC, LRRC42, DYRK2, KLC2, PROM1, SLC18B1, NCOA2, SLC16A4, PRKCZ, SEMA6A, MSC, ATP9B, XPNPEP2, APBB3, PALLD, KITLG, CHST10, LRRC41 , WDR11, LGALS13, GAP43, ACO2, CAPRIN2, AZIN1, MYLK, GUCY1B3, POMGNT1, HSD17B2, GABRB1, PIP5K1B, MSRB2. ,

[0027] The present application also proposes a gene set obtained based on the above screening method, wherein the gene set comprises the following genes: AHNAK, ANG, ANXA4, BCAR3, BCL6, BMPR2, BTD, CALM1, CRH, CSNK2A2, CXXC1, CYP11A1, DCP2, DDR1, DUSP1, DVL1, EBI3, EFHD1, ELF3, ENG, EPS8L1, ERO1L, FAM120A, FGD6, FKBP2, FLT1, GALE, GALNT2, GATA3, GFOD2, GREM2, HEY1, HTRA1, IL1RAP, ILVBL, INHA, INHBA, ITGA5, JUNB, KIAA19 19, KIF2A, KLF4, KRT19, LEP, LRRC1, MAN1C1, MBD2, MTMR4, NDRG1, NEBL, NEK11, OCRL, PAM, PAPPA2, PDLIM2, PHYHIP, PIK3AP1, PJA1, PPL, PROCR, PRRG1, PVRL4, RA B25, RASSF7, SCARB1, SH3BP5, SH3PXD2A, SIPA1L1, SLC11A2, SLC2A1, SLC35F2, SLC6A8, SLCO2A1, SLCO4A1, SPAG4, SSFA2, SYDE1, TBC1D22A, TNIP2, TPBG, ZMYND8.

[0028] The present application also proposes a diagnostic marker for pre-eclampsia, which comprises at least one of the genes NEU1, NEU2, NEU3, and NEU4.

[0029] Optionally, the pre-eclampsia diagnostic marker comprises the gene NEU1.

[0030] This application also proposes the use of the above-mentioned gene set and pre-eclampsia diagnostic markers in the preparation of products for diagnosing pre-eclampsia.

[0031] This application also proposes a method for constructing a disease animal model, comprising the following steps:

[0032] Constructing datasets of gene expression profiles of diseased and non-disease samples from humans and animals respectively; co-clustering the datasets of gene expression profiles of diseased and non-disease samples from humans and animals by coupled non-negative matrix factorization to obtain coupled gene clusters in the datasets of humans and animals;

[0033] Determining transferable genes between humans and animals based on the coupled gene cluster, the screening results of the above-mentioned disease-related gene screening method, and the gene expression levels of animals;

[0034] The disease animal model is constructed based on the transferable gene.

[0035] Optionally, the animal is a mouse.

[0036] This application also proposes the use of the disease animal model obtained by the above-mentioned method for constructing the disease animal model in verifying the effective treatment identification of pre-eclampsia.

[0037] The technical solution of this application has the following advantages:

[0038] This application has begun to combine known regulatory network information with whole-genome expression data through a network model of genotype-phenotype relationship. Through the network mathematical model, the causal relationship from gene to gene product to phenotype can be quantified, and the causal influence of single or polymorphic genes can be simulated.

[0039] This application proposes that the NEU gene can be used as a diagnostic marker for preeclampsia, and proposes the feasibility of a new preeclampsia model based on dysfunction of the key vascular protein NEU1 (an important protein involved in the development of new blood vessels). In the future, this model can be used to test patients and develop new therapies targeting these important mechanisms of preeclampsia. This application determined that NEU1 is transferable between human and mouse models based on the method of coupled non-negative matrix decomposition. Based on this gene, a reliable preclinical model can be constructed to replicate severe human preeclampsia. By combining network and quantitative gene expression variation with phenotypic modeling, it was found that key genes on the network are more likely to contribute to disease or adaptive progression. These genes for severe preeclampsia are recommended as potential druggable targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Interventions to reduce the risk of preeclampsia.

[0041] Figure 2 shows an example of GLM analysis of FLT1 gene and PE classification phenotype;

[0042] Figure 3 shows the cluster analysis results of 844 genes;

[0043] Figure 4 shows the results of IPA and KEGG analysis;

[0044] FIG5 is a translation model for gene expression fitting from mice to humans;

[0045] Figure 6 shows the co-cluster relationship between human and mouse gene expression;

[0046] FIG7 shows the structure of the vascular endothelial glycocalyx (VEGLX);

[0047] Figure 8 is a complete image of the vascular endothelial glycocalyx;

[0048] Figure 9 is the clustering result diagram of the scoring matrix S2. DETAILED DESCRIPTION

[0049] The following examples are provided to further better understand the present application, but are not limited to the best implementation mode described herein, and do not limit the content and protection scope of the present application. Any product identical or similar to the present application obtained by anyone under the inspiration of the present application or by combining the features of the present application with other prior arts shall fall within the scope of protection of the present application.

[0050] If no specific experimental steps or conditions are specified in the examples, the conventional experimental steps or conditions described in the literature in this field can be used. If the manufacturer of the reagents or instruments is not specified, they are all commercially available conventional reagents.

[0051] Example 1

[0052] This embodiment provides a method for screening disease-related genes, comprising the following steps:

[0053] Step 1) constructing a dataset containing gene expression profiles of diseased and non-diseased samples;

[0054] Analysis of gene expression changes holds great potential for dissecting the gene regulatory networks that control important biological processes. Network models of genotype-phenotype relationships have begun to combine known regulatory network information with genome-wide expression data with the goal of predicting an organism's phenotype. A phenotype, in this context, is a "high-dimensional" entity: a combination of clinical phenotypes and transcriptional readouts associated with a specific combination of gene expression. Network mathematical models can quantify the causal relationships from gene to gene product to phenotype. They can simulate the causal effects of single or multiple genes, suggesting that key gene expression variations may be more likely to contribute to complex clinical phenotypic variation. To identify genes involved in severe preeclampsia, 263 preeclampsia patients were collected and linked to a mouse model of preeclampsia. By integrating gene expression variation across patients, including fine-grained and genome-wide RNA profiling, with clinical phenotypes, and integrating network and quantitative gene expression variation with phenotypic modeling, key genes on the network were identified as more likely to contribute to disease or adaptive progression. These genes implicated in severe preeclampsia are recommended as potential druggable targets.

[0055] Four datasets (3 human and 1 mouse) from the Gene Expression Omnibus database were used. The contents of these datasets are as follows:

[0056] GSE75010 contains gene expression profiles of 157 human placental tissue samples. Of these 157 samples, 80 were preeclampsia (PE) samples and 77 were non-PE samples. In addition to gene expression data, the dataset also contains 29 clinical phenotypes such as diagnosis, mean uterine PI, mean umbilical cord PI, maximum systolic blood pressure, maximum diastolic blood pressure, proteinuria pattern, HELLP diagnosis, and gestational age.

[0057] GSE25906 contains expression profiles of 60 human placenta samples, 23 of which are preeclampsia samples and 37 are control samples. This dataset contains four phenotypic features, such as sex, classification, gestational age, and induction of labor.

[0058] GSE14722 contains 46 human basal plate biopsy samples, 22 preterm births and 24 severe preeclampsia samples. This dataset does not contain any clinical phenotypes.

[0059] GSE116751 contains expression profiles of 10 pregnant mouse placenta samples, 4 of which were saline-perfused (control) samples and 6 were AVP-perfused (preeclampsia) samples.

[0060] Step 2) establishing a linear model between each gene expression data in the gene expression profile and whether the disease occurs based on the data set in step 1);

[0061] A general linear model (GLM) was used to associate each gene with the clinical phenotype. i ) gene g i As a predictor variable, the phenotypic characteristics pf j As the response variable, a general linear model is established as shown in the following formula (1):

[0062] If there are N genes in the dataset, for each phenotypic feature, N such models are generated. All N slope values ​​are scaled to the range [-1, 1]. Let sb i,j .express Therefore, sb i,j Can be used as gene g i Phenotype pf j score.

[0063] In Figure 6, we can see a scatter plot of FLT1 gene expression and sample diagnosis (PE / non-PE) for the GSE75010 dataset. The diagnosis phenotype can have two values: non-PE = 1 and PE = 2. The GLM constructed on this gene and diagnosis phenotype has the following form: y = -2.558 + 0.359x (2)

[0064] Step 3) According to the slope of the linear model of each gene obtained in step 2), the differentially expressed genes are screened out, and the differentially expressed genes are the disease-related genes

[0065] Taking formula (2) as an example, where x represents gene expression, y represents diagnosis, and 0.359 represents the slope term for the gene. Expanding this slope value to the range of [-1, 1] yields a score of 0.6590. In other words, the phenotypic diagnostic score for the FLT1 gene is 0.6590.

[0066] When the phenotype is "diagnosis" and the available values ​​are pre-eclampsia (PE) and non-pre-eclampsia (non-PE), a threshold θ is used to filter genes. i,j |>θ then gene g i The threshold θ is set based on the statistical p-value of the model. When θ is equal to 0.13, the corresponding statistical p-value is 0.05; when θ is equal to 0.35, the corresponding p-value is 0.0001.

[0067] In order to further understand the relationship between differentially expressed genes and clinical phenotypes, we first construct a scoring matrix S1 F×DWhere F is the number of clinical phenotypes and D is the number of differentially expressed genes. Thus, the matrix S1 F×D Each entry in is the score of the gene in that column for the phenotype of that row. F×D Perform dichotomy on the following. The dichotomy results show which phenotypes are grouped together and their associations with genes.

[0068] Calculate the differentially expressed genes of all three human datasets (GSE75010, GSE25906, and GSE14722). Among them, a gene set containing 844 genes was selected from the GSE75010 dataset:

[0069] <h2 style=";text-align:left;direction:ltr">ANG,C7orf49,DONSON,SLC38A10,DSCR4,HERPUD1,PDLIM2,HSPA2,LACTB,A DAM17,ANO8,GNPTG,DDIT4L,TMCC3,PMEL,PGRMC1,VASN,PXN,PGK1,FGD6,BC AP31,KRT15,SPOCK2,VWA1,MT1F,BIN2,DDR1,RPS19BP1,PJA1,DAB2IP,KIA A1217,RANBP2,CD9,ISM2,CCDC127,TMEM141,DDX58,PFKFB3,NEBL,HPCA,EP HB3,MED12L,SEMA3G,REEP4,HSPG2,CCDC115,GALNT18,LRRC32,GRHL3,KRT 7,KLHDC7A,METTL6,FABP4,SH3BP2,TNNI2,LAIR2,ARSK,HRASLS2,LAMC2,FA M219A,GPER,GREB1L,CLDN5,MIDN,DNAJB2,ITGB6,MAPK8IP3,FPGS,SIPA1L 1,VEGFC,MAP3K8,OXGR1,SPICE1,PTPN12,ADRA2C,RIOK1,DLX5,SHANK2,ATP 6V0E1,JUNB,NANS,LIMK1,GFOD2,ZFP36L1,CKMT2,TPM1,SFR1,STC2,PTGES ,ATG16L2,FBLIM1,UBR1,SLC2A5,CDK19,PWWP2B,SSR1,PPARG,CEBPA,DEGS1 ,PFKL,TCF7L2,TNR,ERP44,GSE1,DCP2,ABCA12,GPR32,IRF9,TNIP2,ZBTB2 ,FOXJ3,CA4,JUP,CAV3,FGF22,PDIA5,KCTD11,FAM8A1,BCAR3,AP3S1,IFI44 L,STK11IP,DVL1,LBH,SNX9,FN1,NLRP2,MTMR4,TMEM179B,VIT,ANKRD23,A NXA4,ELMO3,PFKP,SSFA2,SLC11A2,MAP3K9,B9D2,POLR2H,INSIG1,FOSL2,D OCK5,UBASH3B,SLC35F2,SYN3,MAPK11,IL1R2,DGKZ,HMOX1,ELF3,SNX33,LY 6D,AHNAK,GATA3,MAN1C1,ABCA7,CXXC1,KCTD9,AVL9,ISLR2,DNAH11,PREP,<h2 style=";text-align:left;direction:ltr">JPH1,ASTE1,IGSF3,CLDN9,NARF,SH3RF1,UPK1B,BMPR2,PDK1,BCAT2,SWAP70,SERPINE1,SLC 2A1,EBI3,PMEPA1,TRIP10,SLC28A1,AMOTL2,MORF4L2,RHOB,SLC38A3,TIMP3,NLRP7,CITED2, C5orf30,PFKFB4,SLC2A3,BASP1,FAM184A,HPCAL1,RIMKLA,FBXO22,PLAC8,EPHA1,KLF4,ANK RD13B,CYP11A1,STX1A,MFAP5,GLA,ELF4,SUV39H1,SLC1A6,EFHD1,HN1L,FAM156A,P2RY2,RAB 3D,IRF2BPL,CSNK2A2,RHOF,OAS3,TGM1,HK1,STK24,RASSF7,CCRN4L,KIF17,C14orf37,CCDC 113,MYCN,SLCO4A1,AQP1,RARRES3,LYN,MED13L,BTNL9,MBD2,ASCL2,ALPK3,KIT,PRR15,FAM4 3A,AGRN,CRKL,PAM,RHOU,MIIP,PPP1R16B,FOXJ2,PTPRU,FKBP2,MMP14,PLEC,CRH,IL1RAP,TM EM59,PADI1,VTCN1,GLIS3,PLBD1,PNCK,BAD,RNF24,SUN2,SPTLC3,ING1,CPOX,ARNT2,GREM2, UGP2,DNAJC1,ACSF2,HEY1,USP18,EFNA1,IGSF8,FRMD4A,MAML2,SLC16A1,KIAA1919,HPS3,PPP1R3B,TNFAIP8,RA B25,MIF,FKBP8,TREM1,FAM120AOS,PPL,ITGA5,HMGN3,EPS8L1,RASSF8,F11R,C4orf3,KIF2A,MXI1,MTUS1,GALE,G APDH, RUSC2, HCAR2, TPI1, LTBP1, GPX3, KLF3, TES, JPH2, TMEM66, SLC2A14, CDO1, NEK6, PARD3, RASGRF2, HMGB3, SEC23A, TMEM184A, GSN, RAB8B, ILVBL, SLC16A3, MOB3A, SERPINB9, EDEM2, SFRP1, PPP1R12C, TRIM24, ANKRD37, ARID3A<h2 style=";text-align:left;direction:ltr">LEPROTL1,KDM3A,CSNK1E,FHL2,CCSAP,NEURL,S100A6,CXCR6,FUT11,DSP, EAF1,GALK1,FAM120A,ERRFI1,VWCE,ALDOA,LRRC8A,AK4,ZMYND8,EZR,TAC C1,NTRK2,ADAM9,ADORA2A,HTRA1,GUCA2A,GPR17,ARRDC3,EGLN3,PRKAB2, SERPINA3,PRSS22,EGLN1,C5orf46,MID1,GPT2,PTBP3,LRRC1,NRIP1,PAPPA 2,CCBP2,ARHGEF4,MGAT3,AIF1L,SLCO2A1,KRT17,NOG,PLOD2,DUSP1,PKM, ASPHD2,PLIN2,CYP2J2,QPCT,HILPDA,RDH13,NEK11,CORO2A,GRHL1,SCARB1 GALNT2, DNM2, PIK3AP1, SYDE1, PRRG1, PIK3R1, LIMCH1, UBE2Q2, GDPD3, HK2, MLL5, TLE6, GPR146, BTD, PHYHIP, MYOF, PRKD3, ZCCHC2, CALM1, SIGLEC6, B CL6, LDHA, RASEF, HEXB, ENG, INHA, INHBA, MME, PVRL4, BACH1, PLA2G16, TSC22D1, TRIM14, TBC1D22A, CCBL1, WDR60, LIMD1, NPNT, TMEM45A, SEMA4C, IPM K,ERGIC1,MTSS1L,BHLHE40,SLC6A8,RFK,CST6,DIO2,C12orf75,PROCR,PI K3CB,ERO1L,LEP,SFXN3,SH3BP5,MYO7B,MICAL3,KRT19,HTRA4,TPBG,SPAG4 ,EFNB1,SASH1,OCRL,SMARCA1,P4HA1,NDRG1,SH3PXD2A,COL17A1,FSTL3,F LNB,FLT1,ELMO1,TRPV6,EPHB6,HSD17B1,ARL4C,TMEM139,PKD2,TRIM40,SE MA3D,PDE7B,SLC20A1,HIST1H1T,ANXA6,BTG2,FBN2,ASAH1,SLC25A27,RAB 15,ZC2HC1A,TMEM136,KLRD1,TMEM168,CAP2,HSD11B2,SELE,STON2,SCN9A,<h2 style=";text-align:left;direction:ltr">ABCD3, PISD, PPAP2B, F13A1, CYTH3, OLFML3, FAM101B, SOD1, CSHL1, LRRC4, SNX27, CYBASC3, SOX5, ABAT, ADAMTSL1, C18orf8, LEPREL1, LRIG3, B3GNT2 ABCG2,GRIK2,DISP1,CNTN5,GABARAPL1,CATSPERB,ARPC5,PLCXD2,SLC16A 10,DAB2,TKT,CD28,MORC4,EGFL6,CADM3,KCTD3,SLCO2B1,SYNPO2L,SLC23A 2,PXK,SLC9A9,PDE5A,IPO13,GCLM,OPRK1,ATP8B4,DNMT3A,LAMA2,IL12RB 2,GSTA3,C12orf39,RNF114,MTSS1,TMEM133,CDYL2,MOB3C,CLDN16,C10or f10, ECM1, ACSS1, NPC1, IGF1, SLAMF1, ATP1A1, HSD3B2, NUFIP2, LRRN3, ADAP2, MAF, LSS, EML2, POSTN, RAPH1, SPATA6, LAMC1, APLN, KL, ANK3, TRK2, WWC 1,CPEB4,MAPK8,GNG2,GOT1,FAM20B,ZNF555,AKIP1,SPON1,TLR7,FAM84B, PRKCE,KCNIP3,RCAN1,HDAC9,RALYL,THSD4,ESRRG,HNRNPA0,IL36RN,CLIC 5,VNN3,ST5,ACO1,LGI2,LCP1,CYBB,TXNRD1,KCNIP1,SH3BGRL2,TRAF3,ZN F554,MRS2,KLHL3,ADCY7,RAD54L2,HPGDS,PLEKHO2,UNC5C,HLF,RAP1GAP2, UBALD2,CD200R1,ATRNL1,ENPP1,LGALS14,NFIA,ZNF682,PSEN2,DOCK8,C5 ,SIK3,MKNK1,FES,BMX,SPRED1,ZBTB39,FOXRED2,ZMYND11,GPR34,TNFRSF1 1A,UBB,OR51B5,FOLR2,SLC19A3,CLCN5,TGFBR3,ABHD3,PMAIP1,TOM1L1,O R51B4,MRPS30,NFE2L2,SLC2A12,SCG2,CYSLTR1,MYO1A,VCP,OTUB2,NCEH1,<h2 style=";text-align:left;direction:ltr">SLIT2,KIAA0226L,PITPNA,LYPD6,PLAC1,HMGN2,HIST2H2BE,SUPT7L,DCAF11,PCBD1,GRINA,CABLES1,C5orf42,ALDH4A1,RAB 5C,FRZB,SFXN4,CHKA,APMAP,SLC26A2,CERS2,ATXN2,MAP7D2,NEDD4L,ITPR1,ENPP2,DSN1,IQGAP2,FRY,LYVE1,MB21D2,LAMB 2,CARD16,GSTK1,GSK3B,MATN2,MFF,NALCN,PIKFYVE,KIRREL,TBC1D9,PDPN,TMEM55A,ICA1,APBB2,ARHGAP17,PIK3CG,BMP5, SOX6,PSMC5,AFF3,FGF10,GUCY1A3,RBM47,PRKG1,GLUL,ASB2,STRADA,CCND2,FILIP1,ST3GAL4,LANCL2,SLC37A1,PDZRN3,TSP AN33,VAV1,DEPDC5,PER3,WNT2,CLDN1,RUNX1T1,AMPD3,OR51B2,CYP4V2,IDO1,PC,XPO6,ZNF583,ZNF641,TYRO3,CYTH4,NSMA F,NCKAP1L,PTPN1,FRMD6,SRPX,ELK1,ACSS3,PIAS2,KLF12,TM9SF3,STXBP6,BHLHE41,FLRT3,PTPRD,TMEM106C,ZNF610,PHTF2 ,ZEB2,HSPB11,P2RY1,RAB3IL1,CRYAB,ABCA11P,LAMC3,TMPRSS3,FMO1,SLC4A4,BMP4,RBBP4,DUSP4,CXorf57,HYOU1,KLHL42 ,ANGPTL2,PARP6,AGPAT5,WDSUB1,ABCA9,ATP5B,C11orf63,DUS2L,PIGX,ZNF542,DACH1,LRRTM2,LRRC39,SEC22C,LINC00472, TSPAN7,GABRA3,CHN1,ASPA,FXYD3,PRTFDC1,GUCY1A2,ANKS1B,SLC25A6,AIF1,PDHA1,BIN1,PCO LCE,DLG2,VTN,FXN,CTLA4,SLC17A5,APBA1,NIN,TC2N,PTP4A1,TXK,MAD2L2,FURIN,ERCC6,SGCB,OSBPL11, RDH10, LONRF2, TMEM176A, EID1, VAV3, SLC35F6, KCNJ5, CX3CR1, TMEM86A, NPL, MAT2A, HDAC3, GRHPR, SPEF2, ST6GAL1, BNC2, PRKCD, CLTC, LRRC42, DYRK2, KLC2, PROM1, SLC18B1, NCOA2, SLC16A4, PRKCZ, SEMA6A, MSC, ATP9B, XPNPEP2, APBB3, PALLD, KITLG, CHST10, LRRC41, WDR11, LGA LS13, GAP43, ACO2, CAPRIN2, AZIN1, MYLK, GUCY1B3, POMGNT1, HSD17B2, GABRB1, PIP5K1B, MSRB2,

[0070] We selected overlapping genes that were upregulated in severe pre-eclampsia conditions across all three datasets. This resulted in a gene set of 81 genes:

[0071] AHNAK, ANG, ANXA4, BCAR3, BCL6, BMPR2, BTD, CALM1, CRH, CSNK2A2, CXXC1, CYP11A1, DCP2, DDR1, DUSP1, DVL1, EBI3, EFHD1, ELF3, ENG, EPS8L1, ERO1L, FAM120A, FGD6, FKBP2, FLT1, GALE, GALNT2, GATA3, GFOD2, GREM2, HEY1, HTRA1, IL1RAP, ILVBL, INHA, INHBA, ITGA5, JUNB, KIAA1919, KIF2A, KLF4, KRT 19, LEP, LRRC1, MAN1C1, MBD2, MTMR4, NDRG1, NEBL, NEK11, OCRL, PAM, PAPPA2, PDLIM2, PHYHIP, PIK3AP1, PJA1, PPL, PROCR, PRRG1, PVRL4, RAB25, RASS F7, SCARB1, SH3BP5, SH3PXD2A, SIPA1L1, SLC11A2, SLC2A1, SLC35F2, SLC6A8, SLCO2A1, SLCO4A1, SPAG4, SSFA2, SYDE1, TBC1D22A, TNIP2, TPBG, ZMYND8

[0072] These genes were identified as potential targets for the treatment of preeclampsia. Pathway analysis was then performed on these genes. Pathway analysis is a very convenient and effective method for understanding the diverse roles and functions of a gene set in different metabolic pathways. For this purpose, Ingenuity Pathway Analysis (IPA) and KEGG programs were used. IPA is a genomics analysis tool used to interpret complex biological and molecular pathway data. Its main functions include information mining based on literature databases, disease-biological function association analysis, gene pathway modeling, and the prediction of new biomarkers and drug targets. KEGG is a tool for understanding cellular molecular functions and systemic relationships, integrating macromolecular information (such as the genome, transcriptome, and proteome) with biochemical pathway and disease information. The differentially expressed genes in GSE75010 are the most diverse and comprehensive gene set, so we used it for pathway analysis. The differentially expressed genes obtained from the GSE75010 dataset were entered into IPA and KEGG, and the results were reviewed. The results are shown in Figure 3.

[0073] It can be seen from the results of pathway analysis that neuraminidase genes (NEU1, NEU2, NEU3, NEU4) are closely related to the clinical phenotype of preeclampsia

[0074] Furthermore, to determine the association between neuraminidase genes (NEU1, NEU2, NEU3, NEU4) and clinical phenotypes of preeclampsia, a new scoring matrix S2 was established with 81 target genes as columns. The selected phenotypes and the four genes NEU1, NEU2, NEU3, and NEU4 from the datasets GSE75010 and GSE25906 constitute the rows. The matrix entries for target genes and phenotypes are the scores obtained using GLM (range [-1, 1]), while the matrix entries for target genes and NEU genes are their correlation coefficients, also in the range [-1, 1]. For example, the target gene g from GSE75010 i The matrix entries for the phenotype maternal age and phenotype maternal age are the g obtained by the GLM i Scores and phenotypes maternal age on dataset GSE75010.

[0075] g from GSE75010 i The matrix entries for NEU1 are the NEU1 gene expression in preeclampsia (PE) samples and the NEU1 gene expression in PE samples in the dataset GSE75010. i The correlation between expression. The bi-clustering of this score matrix S2 will provide clustering of NEU genes with different phenotypes. The clustering results are shown in Figure 9.

[0076] Example 2

[0077] To develop mouse models, it is necessary to identify potential target genes that are transferable from mouse models—that is, genes that exhibit similar patterns and behaviors in preeclampsia mice and human samples. The goal is to identify genes that overlap and exhibit similar patterns in human and mouse models. These genes can then be considered transferable between human and mouse models. This work aims to clarify whether a gene is truly a target for treating preeclampsia through experimental validation in mouse models. To identify transferable gene patterns between human and mouse models, the gene expression data from mouse and human samples were analyzed using the coupled non-negative matrix factorization (CoupledNMF) method to identify similar gene clusters within the datasets.

[0078] Non-negative matrix factorization (NMF) is a method used in signal processing to solve the blind source deconvolution problem and is widely used in cluster analysis, face recognition and text mining. NMF can be defined by the following model: X = WH + Er.....W,H≥0 (3)

[0079] Where X is an n×p data matrix, W and H are n×r and r×p matrices respectively; Er is an error term used to account for inaccurate reconstruction. NMF attempts to explain the main sources of variance in the dataset X with a small number of vectors (r). NMF does not require orthogonality among the components.

[0080] Coupled NMF formulates the clustering problem of the same type of data from two different sources as an optimization problem. Let M and E be two data matrices, M being the score matrix for mice and E being the score matrix for humans. By coupling the M and E matrices and simultaneously maximizing the decomposition of the matrices M and E, this problem is transformed into the following optimization problem to obtain the log-likelihood function L, which is defined as:

[0081] In the above model, L has three tuning parameters: λ1, λ2, and μ. λ1 and λ2 regulate the matrix factorization of human E and mouse M respectively, and μ regulates the interaction between human E and mouse M. is the Frobenius norm, defined as the square root of the sum of the absolute squares of its elements. W1 and H1 are the weight decomposition of the mouse matrix M onto the vectors genes and samples. W2 and H2 are the weight decomposition of the human matrix M onto the vectors genes and samples. tr(*) represents the trace of the square matrix, defined as the sum of the elements on the main diagonal of the square matrix. Essentially, this performs a coupled two-factorization operation to map the consistency features from E and M. The fourth term is the scalar weight factors W1 and W2. Using the particle swarm optimization (PSO) algorithm (see Yongsheng Ding, Lijun Cheng, Witold Pedrycz, Kuangrong Hao. Global Nonlinear Kernel Prediction for Large Data Set With a Particle Swarm-Optimized Interval Support Vector Regression. IEEE Transactions on Neural Networks and Learning Systems, 2015, 26(10)), the log-likelihood function L is solved for the optimal parameters W1 and W2. Using the empirical Bayesian biclustering Bi-EB method (see Yazdanparast A, Li L, Zhang C, Cheng L. Bi-EB: Empirical Bayesian biclustering for pattern recognition in multi-omics data integration across species. Gene (Basel). 2022 Oct 30; 13(11): 1982. doi: 10.3390 / gene 13111982. PMID: 36360219; PMCID: PMC9690013.), the weight factors W1 and W2 biclusters and coupled clusters are plotted. Figure 3 illustrates the concept of coupled clustering on human and mouse datasets. The mouse GSE116751 and human GSE75010 datasets were co-clustered (via CoupledNMF) to obtain coupled gene clusters for the two datasets. Figure 4 shows the results of the co-clustering procedure for mouse data (GSE116751) and human data (GSE75010). We can see three co-clusters highlighted in the figure. Table 2 lists the genes in co-cluster 1. As can be seen from the table, all of them have functions related to angiogenesis. Among the genes belonging to the co-cluster, ANG, GREM2, and NEU1 are all present. Comparing the differential expression of NEU genes in mice and humans, only NEU1 can be reliably recommended as a potential target for mouse models. The other genes were discarded due to their low expression in the mouse dataset.

[0082] Table 2 Functional annotation of cluster 1

[0083] Example 3 Experimental Verification

[0084] A preeclampsia diagnostic model was established using the NEU gene, and the diagnostic accuracy of this model was validated. The results are as follows: The NEU target was disrupted in mice using the drugs L-NAME and zanamivir. L-NAME was administered subcutaneously, while zanamivir was administered intravenously. Five groups of mice were generated using different treatments. The data in the table below confirm the association of the NEU target with preeclampsia and demonstrate that disruption of the NEU gene target can be used to establish a mouse model for preeclampsia research.

[0085] Table 3 Systolic blood pressure data

[0086] Table 4 Diastolic blood pressure data

[0087] Table 5 Average pulsating pressure

[0088] Table 6 24-hour urine protein

[0089] Table 7 Number of pups

[0090] The outer membrane of a cell is covered by a network of highly hydrated carbohydrate fibers. This covering is called a glycocalyx. The general structural components of the glycocalyx are glycoproteins and proteoglycans. The precise biochemical composition and structure of the glycocalyx vary with cell type. Glycoproteins are glycoconjugates in which one or more sugars are covalently linked to the polypeptide backbone of a protein. Proteoglycans are core proteins covalently bound to glycosaminoglycan (GAG) chains, also known as silicic acid polysaccharide proteins. The vascular endothelium is a monolayer of cells lining arteries, capillaries, and veins. It not only serves as a barrier between blood and tissues but also actively controls the degree of vascular relaxation and contraction, and the extravasation of solutes, fluids, macromolecules, hormones, platelets, and blood cells. The structure of the vascular endothelial glycocalyx (VEGLX) is shown in Figure 3.

[0091] VEGLX not only protects the vascular endothelium from turbulent wear caused by blood flow but also plays an important role in various endothelial functions. Healthy vascular endothelium covered with VEGLX contains cytokines, chemokines, receptors, growth factors, gap-binding proteins, extracellular superoxide dismutase (ecSOD), and endothelial nitric oxide synthase (eNOS), as shown in Figure 4.

[0092] Damage to VEGLX has been associated with endothelial dysfunction, severe infection, Kawasaki disease, pregnancy-induced hypertension (preeclampsia), gestational diabetes mellitus, sepsis, acute lung injury, trauma, ischemic cerebral embolism, acute coronary syndrome, and shock.

[0093] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for screening disease-related genes, characterized in that: The steps include: Step 1) constructing a dataset containing gene expression profiles of diseased and non-disease samples; Step 2) establishing a linear model between each gene expression data in the gene expression profile and whether the disease occurs according to the data set in step 1); Step 3) According to the slope of the linear model of each gene obtained in step 2), the genes differentially expressed between the diseased and non-disease samples are screened out, and the differentially expressed genes are the disease-related genes.

2. The method for screening disease-related genes according to claim 1, characterized in that: In the step 1), multiple data sets are constructed; in the step 2), linear models are established for each gene expression data in each data set and whether it is a diseased sample phenotype; in the step 3), genes that are differentially expressed in each data set are screened out according to the slope of the linear model of each gene.

3. The method for screening disease-related genes according to claim 1, characterized in that: The step 2) specifically includes the following steps: using each gene expression data in the gene expression profile as a predictor variable and whether the disease is present as a response variable; and setting the value of the diseased sample phenotype to 1 and the value of the non-disease phenotype to 2, and establishing a linear model.

4. The method for screening disease-related genes according to claim 1, characterized in that: The step 3) specifically includes the following steps: expanding the slope value of the line model of each gene to the range of [-1, 1] to obtain the score value of each gene, comparing the score value with a threshold representing a significant difference, and the gene with a score value greater than the threshold is a differentially expressed gene.

5. The method for screening disease-related genes according to any one of claims 1 to 4, characterized in that: The disease is pre-eclampsia.

6. The method for screening disease-related genes according to claim 5, characterized in that: The data set of step 1) also includes data on clinical phenotypes of pre-eclampsia samples and non-pre-eclampsia samples; The clinical phenotype data include at least one of diagnosis result, average uterine pi, average umbilical cord pi, maximum systolic blood pressure, maximum diastolic blood pressure, proteinuria pattern, HELLP diagnosis, and gestational age; The step 2) also includes establishing a linear model between each gene expression data in the gene expression profile and each clinical phenotype data; After the differentially expressed genes are screened out in step 3), a scoring matrix is ​​established; the rows and columns in the scoring matrix represent the differentially expressed genes and the clinical phenotype types, respectively, and the scores of the linear model obtained in step 2) are filled into the corresponding positions in the scoring matrix; The scoring matrix was subjected to a dichotomous clustering analysis to classify the clinical phenotypes.

7. The method for screening disease-related genes according to claim 1, characterized in that: Pathway analysis was performed on the screened disease-related genes.

8. A gene set, characterized in that The following genes are included: ANG, C7orf49, DONSON, SLC38A10, DSCR4, HERPUD1, PDLIM2, HSPA2, LACTB, ADAM17, ANO8, GNPTG, DDIT4L, TMCC3, PMEL, PGRMC1, VASN, PXN, PGK1,FGD6,BCAP31,KRT15,SPOCK2,VWA1,MT1F,BIN2,DDR1,RPS19BP1,PJA1,DAB2IP,KIAA1217,RANBP2,CD9,ISM2,CCDC127,TMEM141,DDX58,PFKFB3, NEBL,HPCA,EPHB3,MED12L,SEMA3G,REEP4,HSPG2,CCDC115,GALNT18,LRRC32,GRHL3,KRT7,KLHDC7A,METTL6,FABP4,SH3BP2,TNNI2,LAIR2,ARSK,HRAS LS2,LAMC2,FAM219A,GPER,GREB1L,CLDN5,MIDN,DNAJB2,ITGB6,MAPK8IP3,FPGS,SIPA1L1,VEGFC,MAP3K8,OXGR1,SPICE1,PTPN12,ADRA2C,RIOK1,DLX 5, SHANK2, ATP6V0E1, JUNB, NANS, LIMK1, GFOD2, ZFP36L1, CKMT2, TPM1, SFR1, STC2, PTGES, ATG16L2, FBLIM1, UBR1, SLC2A5, CDK19, PWWP2B, SSR1, PPARG ,CEBPA,DEGS1,PFKL,TCF7L2,TNR,ERP44,GSE1,DCP2,ABCA12,GPR32,IRF9,TNIP2,ZBTB2,FOXJ3,CA4,JUP,CAV3,FGF22,PDIA5,KCTD11,FAM8A1,BCAR3 ,AP3S1,IFI44L,STK11IP,DVL1,LBH,SNX9,FN1,NLRP2,MTMR4,TMEM179B,VIT,ANKRD23,ANXA4,ELMO3,PFKP,SSFA2,SLC11A2,MAP3K9,B9D2,POLR2H,IN SIG1,FOSL2,DOCK5,UBASH3B,SLC35F2,SYN3,MAPK11,IL1R2,DGKZ,HMOX1,ELF3,SNX33,LY6D,AHNAK,GATA3,MAN1C1,ABCA7,CXXC1,KCTD9,AVL9,ISLR2,<h2 style=";text-align:left;direction:ltr">DNAH11,PREP,JPH1,ASTE1,IGSF3,CLDN9,NARF,SH3RF1,UPK1B,BMPR2,PDK1, BCAT2,SWAP70,SERPINE1,SLC2A1,EBI3,PMEPA1,TRIP10,SLC28A1,AMOTL2,MO RF4L2,RHOB,SLC38A3,TIMP3,NLRP7,CITED2,C5orf30,PFKFB4,SLC2A3,BASP 1,FAM184A,HPCAL1,RIMKLA,FBXO22,PLAC8,EPHA1,KLF4,ANKRD13B,CYP11A1, STX1A,MFAP5,GLA,ELF4,SUV39H1,SLC1A6,EFHD1,HN1L,FAM156A,P2RY2,RAB3D,IRF2BPL,CSNK2A2,RHOF,OAS3 ,TGM1,HK1,STK24,RASSF7,CCRN4L,KIF17,C14orf37,CCDC113,MYCN,SLCO4A1,AQP1,RARRES3,LYN,MED13L,BT NL9,MBD2,ASCL2,ALPK3,KIT,PRR15,FAM43A,AGRN,CRKL,PAM,RHOU,MIIP,PPP1R16B,FOXJ2,PTPRU,FKBP2,MMP 14,PLEC,CRH,IL1RAP,TMEM59,PADI1,VTCN1,GLIS3,PLBD1,PNCK,BAD,RNF24,SUN2,SPTLC3,ING1,CPOX,ARNT2, GREM2,UGP2,DNAJC1,ACSF2,HEY1,USP18,EFNA1,IGSF8,FRMD4A,MAML2,SLC16A1,KIAA1919,HPS3,PPP1R3B,TN FAIP8,RAB25,MIF,FKBP8,TREM1,FAM120AOS,PPL,ITGA5,HMGN3,EPS8L1,RASSF8,F11R,C4orf3,KIF2A,MXI1,MT US1, GALE, GAPDH, RUSC2, HCAR2, TPI1, LTBP1, GPX3, KLF3, TES, JPH2, TMEM66, SLC2A14, CDO1, NEK6, PARD3, RASGRF2, HMGB3, SEC23A, TMEM184A, GSN, RAB8B, ILVBL, SLC16A3, MOB3A, SERPINB9, EDEM2, SFRP1, PPP1R12C, TRIM24<h2 style=";text-align:left;direction:ltr">ANKRD37,ARID3A,LEPROTL1,KDM3A,CSNK1E,FHL2,CCSAP,NEURL,S100A6,C Xcr6 , fut11 , DSP , EAF1 , Galk1 , Fam120A , Errfi1 , vwce , aldoa , lrc8a , ak4 , ZMYND8, EZR, TACC1, NTRK2, ADAM9, ADORA2A, HTRA1, GUCA2A, GPR17, ARDC3, EGLN3, PRKAB2, SERPINA3, PRSS22, EGLN1, C5orf46, MID1, GPT2, PTBP3, LRR C1,NRIP1,PAPPA2,CCBP2,ARHGEF4,MGAT3,AIF1L,SLCO2A1,KRT17,NOG,PL OD2,DUSP1,PKM,ASPHD2,PLIN2,CYP2J2,QPCT,HILPDA,RDH13,NEK11,CORO2 A,GRHL1,SCARB1,GALNT2,DNM2,PIK3AP1,SYDE1,PRRG1,PIK3R1,LIMCH1,U BE2Q2,GDPD3,HK2,MLL5,TLE6,GPR146,BTD,PHYHIP,MYOF,PRKD3,ZCCHC2,C ALM1, SIGLEC6, BCL6, LDHA, RASEF, HEXB, ENG, INHA, INHBA, MME, PVRL4, BACH1, PLA2G16, TSC22D1, TRIM14, TBC1D22A, CCBL1, WDR60, LIMD1, NPNT, TMEM4 5A, SEMA4C, IPMK, ERGIC1, MTSS1L, BHLHE40, SLC6A8, RFK, CST6, DIO2, C12orf75, PROCR, PIK3CB, ERO1L, LEP, SFXN3, SH3BP5, MYO7B, MICAL3, KRT19, HTR A4,TPBG,SPAG4,EFNB1,SASH1,OCRL,SMARCA1,P4HA1,NDRG1,SH3PXD2A,CO L17A1,FSTL3,FLNB,FLT1,ELMO1,TRPV6,EPHB6,HSD17B1,ARL4C,TMEM139,P KD2,TRIM40,SEMA3D,PDE7B,SLC20A1,HIST1H1T,ANXA6,BTG2,FBN2,ASAH1, SLC25A27,RAB15,ZC2HC1A,TMEM136,KLRD1,TMEM168,CAP2,HSD11B2,SELE,<h2 style=";text-align:left;direction:ltr">STON2, SCN9A, ABCD3, PISD, PPAP2B, F13A1, CYTH3, OLFML3, FAM101B, SOD1, CSHL1, LRRC4, SNX27, CYBASC3, SOX5, ABAT, ADAMTSL1, C18orf8, LEPREL1, L RIG3, B3GNT2, ABCG2, GRIK2, DISP1, CNTN5, GABARAPL1, CATSPERB, ARPC5, PLCXD2, SLC16A10, DAB2, TKT, CD28, MORC4, EGFL6, CADM3, KCTD3, SLCO2B1, SY NPO2L, SLC23A2, PXK, SLC9A9, PDE5A, IPO13, GCLM, OPRK1, ATP8B4, DNMT3A, LAMA2, IL12RB2, GSTA3, C12orf39, RNF114, MTSS1, TMEM133, CDYL2, MOB3C CLDN16,C10orf10,ECM1,ACSS1,NPC1,IGF1,SLAMF1,ATP1A1,HSD3B2,NUFI P2,LRRN3,ADAP2,MAF,LSS,EML2,POSTN,RAPH1,SPATA6,LAMC1,APLN,KL,AN K3,TRAK2,WWC1,CPEB4,MAPK8,GNG2,GOT1,FAM20B,ZNF555,AKIP1,SPON1, TLR7,FAM84B,PRKCE,KCNIP3,RCAN1,HDAC9,RALYL,THSD4,ESRRG,HNRNPA0 ,IL36RN,CLIC5,VNN3,ST5,ACO1,LGI2,LCP1,CYBB,TXNRD1,KCNIP1,SH3BG RL2,TRAF3,ZNF554,MRS2,KLHL3,ADCY7,RAD54L2,HPGDS,PLEKHO2,UNC5C,H LF,RAP1GAP2,UBALD2,CD200R1,ATRNL1,ENPP1,LGALS14,NFIA,ZNF682,PS EN2,DOCK8,C5,SIK3,MKNK1,FES,BMX,SPRED1,ZBTB39,FOXRED2,ZMYND11,G PR34,TNFRSF11A,UBB,OR51B5,FOLR2,SLC19A3,CLCN5,TGFBR3,ABHD3,PMA IP1,TOM1L1,OR51B4,MRPS30,NFE2L2,SLC2A12,SCG2,CYSLTR1,MYO1A,VCP,<h2 style=";text-align:left;direction:ltr">OTUB2, NCEH1, SLIT2, KIAA0226L, PITPNA, LYPD6, PLAC1, HMGN2, HIST2H2BE, SUP7L, DCAF11, PCBD1, GRINA, CABLES1, C5orf42, ALDH4A1, RAB5C, FRZB, SFXN4, CHKA, APMAP, SLC26A2, CERS2, ATXN2, MAP7D2, NEDD4L, ITPR1, ENPP2, DSN1, IQG AP2,FRY,LYVE1,MB21D2,LAMB2,CARD16,GSTK1,GSK3B,MATN2,MFF,NALCN,PIKFYVE,KIRREL,TBC1D9,PDPN,TMEM55A,I CA1,APBB2,ARHGAP17,PIK3CG,BMP5,SOX6,PSMC5,AFF3,FGF10,GUCY1A3,RBM47,PRKG1,GLUL,ASB2,STRADA,CCND2,FI LIP1,ST3GAL4,LANCL2,SLC37A1,PDZRN3,TSPAN33,VAV1,DEPDC5,PER3,WNT2,CLDN1,RUNX1T1,AMPD3,OR51B2,CYP4V2 ,IDO1,PC,XPO6,ZNF583,ZNF641,TYRO3,CYTH4,NSMAF,NCKAP1L,PTPN1,FRMD6,SRPX,ELK1,ACSS3,PIAS2,KLF12,TM9S F3,STXBP6,BHLHE41,FLRT3,PTPRD,TMEM106C,ZNF610,PHTF2,ZEB2,HSPB11,P2RY1,RAB3IL1,CRYAB,ABCA11P,LAMC3, TMPRSS3,FMO1,SLC4A4,BMP4,RBBP4,DUSP4,CXorf57,HYOU1,KLHL42,ANGPTL2,PARP6,AGPAT5,WDSUB1,ABCA9,ATP5B, C11orf63, DUS2L, PIGX, ZNF542, DACH1, LRRTM2, LRRC39, SEC22C, LINC00472, TSPAN7, GABRA3, CHN1, ASPA, FXYD3, PRTFDC1, GUCY1A2, ANKS1B, SLC25A6, AIF1, PDHA1, BIN1, PCOLCE, DLG2, VTN, FXN, CTLA4, SLC17A5, APBA1, NIN, TC2N, PTP4A1, TXK, MAD2L2, FURINERCC6, SGCB, OSBPL11, RDH10, LONRF2, TMEM176A, EID1, VAV3, SLC35F6, KCNJ5, CX3CR1, TMEM86A, NPL, MAT2A, HDAC3, GRHPR, SPEF2, ST6GAL1, BNC2, PRKCD, CLTC, LRRC42, DYRK2, KLC2, PROM1, SLC18B1, NCOA2, SLC16A4, PRKCZ, SEMA6A, MSC, ATP9B, XPNPEP2, APBB3, PALLD, KITLG, CHST10, LRRC41, WDR11, LGALS13, GAP43, ACO2, CAPRIN2, AZIN1, MYLK, GUCY1B3, POMGNT1, HSD17B2, GABRB1, PIP5K1B, MSRB2., 9. A gene set, characterized in that: The gene set includes the following genes: AHNAK, ANG, ANXA4, BCAR3, BCL6, BMPR2, BTD, CALM1, CRH, CSNK2A2, CXXC1, CYP11A1, DCP2, DDR1, DUSP1, DVL1, EBI3, EFHD1, ELF3, ENG, EPS8L1, ERO1L, FAM120A, FGD6, FKBP2, FLT1, GALE, GALNT2, GATA3, GFOD2, GREM2, HEY1, HTRA1, IL1RAP, ILVBL, INHA, INHBA, ITGA5, JUNB, KIAA1919, KIF2A, KLF 4. KRT19, LEP, LRRC1, MAN1C1, MBD2, MTMR4, NDRG1, NEBL, NEK11, OCRL, PAM, PAPPA2, PDLIM2, PHYHIP, PIK3AP1, PJA1, PPL, PROCR, PRRG1, PVRL4, RAB25, RA SSF7, SCARB1, SH3BP5, SH3PXD2A, SIPA1L1, SLC11A2, SLC2A1, SLC35F2, SLC6A8, SLCO2A1, SLCO4A1, SPAG4, SSFA2, SYDE1, TBC1D22A, TNIP2, TPBG, ZMYND8.

10. A diagnostic marker for pre-eclampsia, characterized in that: The pre-eclampsia diagnostic marker comprises at least one of the genes NEU1, NEU2, NEU3, and NEU4.

11. The pre-eclampsia diagnostic marker according to claim 10, characterized in that: The pre-eclampsia diagnostic marker comprises the gene NEU1.

12. Use of the gene set according to claim 8 or 9, or the pre-eclampsia diagnostic marker according to claim 10 or 11 in the preparation of a product for diagnosing pre-eclampsia.

13. A method for constructing a disease animal model, characterized in that: The steps include: Constructing data sets of gene expression profiles of diseased and non-disease samples of humans and animals respectively; co-clustering the data sets of gene expression profiles of diseased and non-disease samples of humans and animals by coupled non-negative matrix decomposition to obtain coupled gene clusters in the human and animal data sets; Determining transferable genes between humans and animals based on the coupled gene cluster, the screening results of the disease-related gene screening method according to any one of claims 1 to 7, and the gene expression level of the animal; The disease animal model is constructed based on the transferable gene.

14. The method for constructing an animal model according to claim 13, characterized in that: The animal is a mouse.

15. The method for constructing an animal model according to claim 13, characterized in that: The transferable gene is NEU1.

16. Use of the disease animal model obtained according to the method for constructing a disease animal model according to claim 13 or 14 in verifying the effective treatment identification of pre-eclampsia.

Images