Data analysis of GC-MS and LC-MS in Envrionmental Science

Data structure of GC/LC-MS full scan mode

GC/LC is used for separation and MS is used for detection in a GC/LC-MS system. The collected data are intensities of certain mass at different retention time. When we perform analysis on certain column in full scan mode, the counts of different mass were collected in each scan. The drwell time for each scan might only last for 500ms or less. Then the next scan begins with a different retention time. Here we could use a matrix to stand for those data. Each column stands for each mass and row stands for the retention time of that scan. Such matrix could be treated as time series data. In this package, we treat such data as matrix type.

For high-resolution MS, building such matrix is tricky. We might need to bin the RAW data to make alignment for different scans into a matrix. Such works could be done by xcms.

Data structure of GC-MS SIM mode

When you perform a selected ions monitor(SIM) mode analysis, only few mass data were collected and each mass would have counts and retention time as a time seris data. In this package, we treat such data as data.frame type.

Data input

You could use getmd to import the mass spectrum data as supported by xcms and get the profile of GC-MS data matrix. mzstep is the bin step for mass:

data <- getmd('data/data1.CDF', mzstep = 0.1)

You could also subset the data by the mass(m/z 100-1000) or retention time range(40-100s) in getmd function:

data <- getmd(data,mzrange=c(100,1000),rtrange=c(40,100))

You could also combined the mass full-scan data with the same range of retention time by cbmd:

data <- cbmd(data1,data2,data3)

Basic visulation of mass spectrum data

You could plot the Total Ion Chromatogram(TIC) for certain RT and mass range.

plottic(data,rt=c(3.1,25),ms=c(100,1000))

You could also plot the mass spectrum for certain RT range. You could use the returned MSP files for NIST search:

plotrtms(data,rt=c(3.1,25),ms=c(100,1000))

The Extracted Ion Chromatogram(EIC) is also support by enviGCMS and the returned data could be analysised for molecular isotopes:

plotmsrt(data,ms=500,rt=c(3.1,25))

You could use plotms or plotmz to show the heatmap or scatter plot for LC/GC-MS data, which is very useful for exploratory data analysis.

plotms(data)
plotmz(data)

You could change the retention time into the temprature if it is a constant speed of temperature rising process. But you need show the temprature range.

plott(data,temp = c(100,320))

Data analysis for influnces from GC-MS

enviGCMS supplied many function for decreasing the noisy during the analysis process. findline could be used for find line of the boundary regression model for noisy. comparems could be used to make a point-to-point data subtraction of two full-scan mass spectrum data. plotgroup could be used convert the data matrix into a 0-1 heatmap according to threshold. plotsub could be used to show the self backgroud subtraction of full-scan data. plotsms shows the RSD of the intensity of full scan data. plothist could be used to find the data distribution of the histgram of the intensities of full scan data.

Data analysis for molecular isotope ratio

Some functions could be used to caculate the molecular isotope ratio. EIC data could be import into GetIntergration and return the infomation of found peaks. Getisotoplogues could be used to caculate the molecular isotope ratio of certain molecular. Some shortcut function such as batch and qbatch could be used to caculate molecular isotope ratio for mutiple and single molecular in EIC data.

Additional functions for xcms package

In environmetnal non-target analysis, some specific functions such as comparision, batch correction and visulization could be useful to handle data pre-processed by xcms as shown below. Besides, isotope extraction for single group of samples with certain mass diff could also be done by getmassdiff function.

• getdata could be used to get the xcmsSet object in one step with optimized methods

• getbgremove could subtract the background group from xcmsSet object with two groups (the first group is background group)

• getupload could get the csv files to be submitted to Metaboanalyst

• gettechrep could get the report for technique replicates

• getbiotechrep could get the report for all of your technique replicates for bio replicated sample in single group

• getgrouprep could get the report for samples with biological and technique replicates in different groups

• gettimegrouprep could get the report for the time series or two factor groups for samples with biological and technique replicates in different groups

• plotmr could plot the scatter plot for xcmsset (or two) objects with threshold

• plote could plot EIC and boxplot for all peaks and return diffreport

• getfeaturest could get the features from t test, with p value, q value, rsd and power restriction

• getfeaturesanova could get the features from anova, with p value, q value, rsd and power restriction

• svacor could perform surrogate variable analysis(SVA) to correct the batch effects

• svapca, svaplot, svapca could be used to visulize the batch effects and their influnces on each peaks

Summary

In general, enviGCMS could be used to explore the data from GC/LC-MS and extract certain patterns in the data with various purposes.