GIPMA

Description

GIPMA

Two dimensional (2D) 1H-13C HSQC has been increasingly applied to metabolomics study because it can greatly improve the resolving capability compared with 1D 1H NMR. However, the preprocessing methods such as peak matching and alignment tools for 2D NMR based metabolomics have lagged behind similar methods for 1D 1H NMR based metabolomics. Here, we developed an efficient peak matching and alignment algorithm (GIPMA) for 2D 1H-13C HSQC based metabolomics, which has been implemented in an open access R-based package called GIPMA (Intelligent peak matching and alignment for metabolomics).

In GIPMA, fully automatic peak-wise matching and alignment do not require any spectrum as initial reference, while chemical shift of each vPC (or spectral feature) is updated and warranted to be statistically more accurate by the intensity-weighted average of the chemical shifts of all peaks in the same vPC. Accurate chemical shifts for each representative spectral feature will facilitate subsequent peak assignment and are essential for correct metabolite identification and result interpretation.

Download GIPMA

Functions

See the R help page for more details!!!

1."CreatSampleId": The purpose of "CreatSampleId"is to generate sample labels for distinguishing the peaks to which the sample belong and for subsequent statistics of vPC.

2.{"Newton_DataOrganize","Sparky_DataOrganize"," Topspin_DataOrganize"}: Organize the layout of CSV data that from Newton,Sparky or Topspin, and integrate it into a data frame (Global peak list)for the followed analysis.

3."PairingCHTol": The initial tolerance ranges of H and C chemical shifts are estimated to be in a predefined generous tolerance range (It is a user defined parameters).

4."SubSampling": In order to expedite the computational efficiency, a set of representative samples were uniformly subsampled from each sample group in case there are too many samples in multiple groups.

5."OptCHTol": Proposed an adaptive method for intelligently selecting the optimum chemical shift tolerance by parallel hill climbing algorithm or partial traversing search in predefined range to maximize the number of vPC among samples.

6."DesignatePC": The intensity-ranked peaks in the global peak list were sequentially designated to specific peak clusters in a self-adaptive way.

7."vPCmatrix": The purpose of "vPCmatrix" is to generate the regularized data matrix of peak intensity (M) with the different spectral features and samples as columns and rows respectively.

8."HCmatrix": In order to facilitate the subsequent identification of differential variables, a matrix containing the accurate chemical shifts of vPC was generated.

Workflow for the Matching and Alignment of 2D HSQC NMR Peaks through GIPMA

For Demonstration Purposes Only!!!

If you want to run the following code, you need to modify the file path and some parameters according to your needs!!!

install package

install.packages("~/GIPMA_0.3.4.zip",repos = NULL)

step1

library(GIPMA)
Sample_1<-CreatSampleId(c("Dd"),7)
Sample_2<-CreatSampleId(c("Dc"),5)
Sample_3<-CreatSampleId(c("Dn"),6)
Sample_4<-CreatSampleId(c("Df"),4)
Sample_5<-CreatSampleId(c("Do"),7)
Sampleid_1<-sort(paste(c(Sample_1,Sample_2,Sample_3,Sample_4,Sample_5),
                       ".1",sep = ""))
Sampleid_2<-sort(paste(c(Sample_1,Sample_2,Sample_3,Sample_4,Sample_5),
                       ".2",sep = ""))
Sampleid_3<-sort(paste(c(Sample_1,Sample_2,Sample_3,Sample_4,Sample_5),
                       ".3",sep = ""))
Sampleid<-c(Sampleid_1,Sampleid_2,Sampleid_3)

step2

path_1<-"E:/Data analysis/raw_all_data.csv"
path_2<-"E:/Data analysis/temp.csv"
path_3<-"E:/Data analysis/All_data.csv"
All_data<-Newton_DataOrganize(path_1,path_2,path_3)

step3

H_Ltol<-0.01;H_Rtol<-0.04;C_Ltol<-0.1;C_Rtol<-0.4
C_H_tol<-PairingCHTol(H_Ltol,H_Rtol,C_Ltol,C_Rtol)

step4

path_4<-"E:/Data analysis/Sampling_data.csv"
Sampling_data<-SubSampling(path_4)

step5

option<-2; H_ppm<-13; C_ppm<-150
Select_C_H<-OptCHTol(option,H_ppm,C_ppm)

step6

path_5<-"E:/Data analysis/order_data.csv"
order_data<-DesignatePC(path_5,H_ppm,C_ppm)

step7

path_6<-"E:/Data analysis/SH_ft2"
path_7<-"E:/Data analysis/Final_data.csv"
Final_data<-vPCmatrix(path_6,path_7)

step8

path_8<-"E:/Data analysis/C_H_data.csv"
C_H_data<-HCmatrix(path_8)

References

Huan Du, Xiu Gu, Jialuo Chen, Caihong Bai, Xiaohui Duan, Kaifeng Hu. GIPMA: Global Intensity-guided Peak Matching and Alignment for 2D 1H-13C HSQC based metabolomics. Analytical Chemistry, 2023,95(6):3195-3203.

Contact

If you have any questions, please do not hesitate to contact us at duhuan@stu.cdutcm.edu.cn or kaifenghu@cdutcm.edu.cn