specmine (version 2.0.3)

nmr_identification: Function to do NMR metabolite identification

Description

This function performs metabolite identification on a dataset of nmr peaks, finding the top 5 reference metabolites that matched with each cluster of variables formed.

Usage

nmr_identification(dataset, ppm.tol=0.03,
                   clust.method='pearson', clust.treshold=NULL, clust.peaks.min=2,
                   clust.maxPeaks=40, clust.nTop=5,
                   freq=500, nucl="1H", solv=NULL, pH=NULL, temp=NULL)

Arguments

dataset

List representing the dataset from an nmr peaks metabolomics experiment.

ppm.tol

ppm tolerance when matching reference peaks to the dataset peaks.

clust.method

Correlation method to use in the formation of clusters. Defaults to "pearson".

clust.treshold

Minimum correlation between variables to form clusters. If not given, the function calculates the optimum value (value that leads to the greater number of clusters).

clust.peaks.min

Minimum number of variables in each cluster. Only the clusters with at least clust.peaks.min variables will be considered.

clust.nTop

Number of top metabolites with greater score to show for each cluster.

clust.maxPeaks

Maximum number of peaks that a cluster can have, while searching for the best correlation value. Defaults to 40, the original value where the code was adapated from. Can also be NULL and this value will be the number of peaks of the larger cluster.

freq

Frequency of reference spectra. See documentation on choose_nmr_references function for further details.

nucl

Atomic nuclei of reference spectra, either "1H" or "13C".

solv

Solvent. See documentation on choose_nmr_references function for further details.

pH

pH. See documentation on choose_nmr_references function for further details.

temp

Temperature. See documentation on choose_nmr_references function for further details.

Value

List with of the results for each cluster. For each cluster:

cluster.peaks

The peaks of the cluster.

summary

Scores of each top 5 reference metabolite matched with the cluster.

metabolites.matched

List with full results for each top 5 reference metabolite matched.

For each reference metabolite matched:

score

Jaccard index score.

matched_peaks_ref

Matched peaks of the reference spectra.

matched_peaks_clust

Matched peaks of the cluster.

reference_peaks

All the reference spectra peaks.

Examples

# NOT RUN {
#data("propolis")
#propolis_mv=missingvalues_imputation(propolis)
#id_res=nmr_identification(propolis_mv, ppm.tol=0.03,
#                   clust.method='pearson', clust.maxPeaks=NULL, clust.nTop=10,
#                   freq=500, nucl="1H")
# }