Non-target Analysis for Wastewater Treatment Evaluation
Ricardo Cunha
cunha@iuta.de13 June, 2025
Source:vignettes/articles/demo_nts.Rmd
demo_nts.RmdIntroduction
In this article we demonstrate how StreamFind can be used to evaluate ozonation of secondary wastewater effluent (i.e., effluent of the aerated biological treatment) using mass spectrometry (MS). A set of 18 mzML files are used, representing blank, influent and effluent measurements in triplicate for both positive and negative ionization modes.
basename(files) [1] "01_tof_ww_is_neg_blank-r001.mzML"
[2] "01_tof_ww_is_neg_blank-r002.mzML"
[3] "01_tof_ww_is_neg_blank-r003.mzML"
[4] "01_tof_ww_is_pos_blank-r001.mzML"
[5] "01_tof_ww_is_pos_blank-r002.mzML"
[6] "01_tof_ww_is_pos_blank-r003.mzML"
[7] "02_tof_ww_is_neg_influent-r001.mzML"
[8] "02_tof_ww_is_neg_influent-r002.mzML"
[9] "02_tof_ww_is_neg_influent-r003.mzML"
[10] "02_tof_ww_is_pos_influent-r001.mzML"
[11] "02_tof_ww_is_pos_influent-r002.mzML"
[12] "02_tof_ww_is_pos_influent-r003.mzML"
[13] "03_tof_ww_is_neg_o3sw_effluent-r001.mzML"
[14] "03_tof_ww_is_neg_o3sw_effluent-r002.mzML"
[15] "03_tof_ww_is_neg_o3sw_effluent-r003.mzML"
[16] "03_tof_ww_is_pos_o3sw_effluent-r001.mzML"
[17] "03_tof_ww_is_pos_o3sw_effluent-r002.mzML"
[18] "03_tof_ww_is_pos_o3sw_effluent-r003.mzML"The showcase will use the StreamFind MassSpecEngine, which encapsulates all tools required for parsing, storing, processing and visualizing MS data. Note that not all methods/functions will be shown, as the demonstration focuses of the workflow to assess wastewater treatment. Other processing methods for MS data are available in the StreamFind package and can be found in the StreamFind reference documentation.
MassSpecEngine
The R6 MassSpecEngine
class object is created using MassSpecEngine$new(), as
shown below. The analyses argument can be used to add the
set of ms files directly. In this demonstration, we use mzML
files. However, the original ms vendor files can also be used, but they
will be converted to mzML format using msConvert from
ProteoWizard
in the background.
# Creates a MassSpecEngine from mzML files
ms <- MassSpecEngine$new(
metadata = list(name = "Wastewater NTA"),
analyses = files
)
# MassSpecEngine class hierarchy
class(ms)[1] "MassSpecEngine" "CoreEngine" "R6"
# Number of analyses
length(ms$Analyses)[1] 18Metadata
Project metadata (e.g., name, author and description) can be added to
the MassSpecEngine$Metadata as a named list object as shown
below. The elements of the list can be anything but must have length
one.
# Adds metadata to the MassSpecEngine
ms$Metadata <- list(
name = "Wastewater Ozonation Showcase",
author = "Ricardo Cunha",
description = "Demonstration project"
)
# Gets and shows the metadata
show(ms$Metadata)name: Wastewater Ozonation Showcase
author: Ricardo Cunha
description: Demonstration project
date: 2025-06-13 12:14:59.534834
file: NA
# Gets the MassSpecEngine date
ms$Metadata@entries$date[1] "2025-06-13 12:14:59 CEST"Replicates and blanks
The analysis replicate names and the associated blank replicate name
can be amended in the MassSpecEngine, as shown below.
Alternatively, a data.frame with column names
file, replicate and blank could be added as
the analyses argument in
MassSpecEngine$new(analyses = files) to have directly the
replicate and blank replicate names assigned (more details here).
# Character vector with analysis replicate names
rpls <- c(
rep("blank_neg", 3),
rep("blank_pos", 3),
rep("influent_neg", 3),
rep("influent_pos", 3),
rep("effluent_neg", 3),
rep("effluent_pos", 3)
)
# Character vector with associated blank replicate names
# Note that the order should match the respective replicate
blks <- c(
rep("blank_neg", 3),
rep("blank_pos", 3),
rep("blank_neg", 3),
rep("blank_pos", 3),
rep("blank_neg", 3),
rep("blank_pos", 3)
)
# Amends replicate and blank names
ms$add_replicate_names(rpls)
ms$add_blank_names(blks)
# Replicates and blanks were amended
ms$Analyses$info[, c(1:3, 5)] analysis replicate blank polarity
<char> <char> <char> <char>
1: 01_tof_ww_is_neg_blank-r001 blank_neg blank_neg negative
2: 01_tof_ww_is_neg_blank-r002 blank_neg blank_neg negative
3: 01_tof_ww_is_neg_blank-r003 blank_neg blank_neg negative
4: 01_tof_ww_is_pos_blank-r001 blank_pos blank_pos positive
5: 01_tof_ww_is_pos_blank-r002 blank_pos blank_pos positive
6: 01_tof_ww_is_pos_blank-r003 blank_pos blank_pos positive
7: 02_tof_ww_is_neg_influent-r001 influent_neg blank_neg negative
8: 02_tof_ww_is_neg_influent-r002 influent_neg blank_neg negative
9: 02_tof_ww_is_neg_influent-r003 influent_neg blank_neg negative
10: 02_tof_ww_is_pos_influent-r001 influent_pos blank_pos positive
11: 02_tof_ww_is_pos_influent-r002 influent_pos blank_pos positive
12: 02_tof_ww_is_pos_influent-r003 influent_pos blank_pos positive
13: 03_tof_ww_is_neg_o3sw_effluent-r001 effluent_neg blank_neg negative
14: 03_tof_ww_is_neg_o3sw_effluent-r002 effluent_neg blank_neg negative
15: 03_tof_ww_is_neg_o3sw_effluent-r003 effluent_neg blank_neg negative
16: 03_tof_ww_is_pos_o3sw_effluent-r001 effluent_pos blank_pos positive
17: 03_tof_ww_is_pos_o3sw_effluent-r002 effluent_pos blank_pos positive
18: 03_tof_ww_is_pos_o3sw_effluent-r003 effluent_pos blank_pos positiveProcessingStep
Data processing is performed by steps according to
ProcessingStep objects. S7 ProcessingStep class
objects are obtained via the respective [Engine type]Method_[method name]_[algorithm name]
constructor functions, attributing the respective subclass. Below we
obtain the ProcessingStep for the method
FindFeatures using the algorithm openms. The
parameters can be changed via the constructor arguments. Documentation
for each ProcessingStep subclass can be found in the StreamFind
reference documentation.
# Gets ProcessingStep for finding features using the openms algorithm
ffs <- MassSpecMethod_FindFeatures_openms(
noiseThrInt = 1000,
chromSNR = 3,
chromFWHM = 7,
mzPPM = 15,
reEstimateMTSD = TRUE,
traceTermCriterion = "sample_rate",
traceTermOutliers = 5,
minSampleRate = 1,
minTraceLength = 4,
maxTraceLength = 70,
widthFiltering = "fixed",
minFWHM = 4,
maxFWHM = 35,
traceSNRFiltering = TRUE,
localRTRange = 0,
localMZRange = 0,
isotopeFilteringModel = "none",
MZScoring13C = FALSE,
useSmoothedInts = FALSE,
intSearchRTWindow = 3,
useFFMIntensities = FALSE,
verbose = FALSE
)
# Prints in console the details of the ProcessingStep
show(ffs)
StreamFind::MassSpecMethod_FindFeatures_openms
data_type MassSpec
method FindFeatures
required NA
algorithm openms
version 0.2.0
software openms
developer Oliver Kohlbacher
contact oliver.kohlbacher@uni-tuebingen.de
link https://openms.de/
doi https://doi.org/10.1038/nmeth.3959
parameters:
- noiseThrInt 1000
- chromSNR 3
- chromFWHM 7
- mzPPM 15
- reEstimateMTSD TRUE
- traceTermCriterion sample_rate
- traceTermOutliers 5
- minSampleRate 1
- minTraceLength 4
- maxTraceLength 70
- widthFiltering fixed
- minFWHM 4
- maxFWHM 35
- traceSNRFiltering TRUE
- localRTRange 0
- localMZRange 0
- isotopeFilteringModel none
- MZScoring13C FALSE
- useSmoothedInts FALSE
- intSearchRTWindow 3
- useFFMIntensities FALSE
- verbose FALSE
# Creates an ordered list with all processing steps for the MS data
workflow <- list(
# Find features using the openms algorithm, created above
ffs,
# Annotation of natural isotopes and adducts
MassSpecMethod_AnnotateFeatures_StreamFind(
rtWindowAlignment = 0.3,
maxIsotopes = 8,
maxCharge = 2,
maxGaps = 1
),
# Excludes annotated isotopes and adducts
MassSpecMethod_FilterFeatures_StreamFind(
excludeIsotopes = TRUE,
excludeAdducts = TRUE
),
# Grouping features across analyses
MassSpecMethod_GroupFeatures_openms(
rtalign = FALSE,
QT = FALSE,
maxAlignRT = 5,
maxAlignMZ = 0.008,
maxGroupRT = 5,
maxGroupMZ = 0.008,
verbose = FALSE
),
# Filter feature groups with maximum intensity below 5000 counts
MassSpecMethod_FilterFeatures_StreamFind(
minIntensity = 3000
),
# Fill features with missing data
# Reduces false negatives
MassSpecMethod_FillFeatures_StreamFind(
withinReplicate = FALSE,
rtExpand = 2,
mzExpand = 0.0005,
minTracesIntensity = 1000,
minNumberTraces = 6,
baseCut = 0.3,
minSignalToNoiseRatio = 3,
minGaussianFit = 0.2
),
# Calculate quality metrics for each feature
MassSpecMethod_CalculateFeaturesQuality_StreamFind(
filtered = FALSE,
rtExpand = 2,
mzExpand = 0.0005,
minTracesIntensity = 1000,
minNumberTraces = 6,
baseCut = 0
),
# Filter features based on minimum signal-to-noise ratio (s/n)
# The s/n is calculated using the CalculateFeaturesQuality method
MassSpecMethod_FilterFeatures_StreamFind(
minSnRatio = 5
),
# Filter features using other parameters via the patRoon package
MassSpecMethod_FilterFeatures_patRoon(
maxReplicateIntRSD = 40,
blankThreshold = 5,
absMinReplicateAbundance = 3
),
# Finds internal standards in the MS data
# db_is is a data.table with the
# name, mass and expected retention time of
# spiked internal standards, as shown below
MassSpecMethod_FindInternalStandards_StreamFind(
database = db_is,
ppm = 8,
sec = 10
),
# Corrects matrix suppression using the TiChri method from
# 10.1021/acs.analchem.1c00357 to better compare influent and effluent
MassSpecMethod_CorrectMatrixSuppression_TiChri(
mpRtWindow = 10,
istdAssignment = "range",
istdRtWindow = 50,
istdN = 2
),
# Loads MS1 for features not filtered
MassSpecMethod_LoadFeaturesMS1_StreamFind(
filtered = FALSE
),
# Loads MS2 for features not filtered
MassSpecMethod_LoadFeaturesMS2_StreamFind(
filtered = FALSE
),
# Loads feature extracted ion chromatograms (EIC)
MassSpecMethod_LoadFeaturesEIC_StreamFind(
filtered = FALSE
),
# Performs suspect screening using the StreamFind algorithm
# db_with_ms2 is a database with suspect chemical standards
# includes MS2 data (i.e., fragmentation pattern) from standards
MassSpecMethod_SuspectScreening_StreamFind(
database = db_with_ms2,
ppm = 10,
sec = 15,
ppmMS2 = 10,
minFragments = 3
)
)Then, the list can be added to the MassSpecEngine. Note that the order will matter when the workflow is applied!
# Conversion of list to Workflow object with validation
workflow <- Workflow(workflow)
# Adds the workflow to the engine. The order matters!
ms$Workflow <- workflow
# Printing the data processing workflow
show(ms$Workflow)1: FindFeatures (openms)
2: AnnotateFeatures (StreamFind)
3: FilterFeatures (StreamFind)
4: GroupFeatures (openms)
5: FilterFeatures (StreamFind)
6: FillFeatures (StreamFind)
7: CalculateFeaturesQuality (StreamFind)
8: FilterFeatures (StreamFind)
9: FilterFeatures (patRoon)
10: FindInternalStandards (StreamFind)
11: CorrectMatrixSuppression (TiChri)
12: LoadFeaturesMS1 (StreamFind)
13: LoadFeaturesMS2 (StreamFind)
14: LoadFeaturesEIC (StreamFind)
15: SuspectScreening (StreamFind)
run_workflow()
The Workflow can be applied by run_workflow(),
as demonstrated below. Note that with run_workflow(), the
processing modules are applied with the same order as they were
added.
# Runs all ProcessingStep added
ms$run_workflow()Results
The created features and feature groups can be inspected as
data.table objects or plotted by dedicated methods in the
MassSpecEngine or methods of the created Results class.
Internally, the MassSpecEngine stores the results in the
NonTargetAnalysisResults object, which can be accessed with
MassSpecEngine$NonTargetAnalysisResults. Yet, the engine
interface is recommended for accessing the results. However, the
NonTargetAnalysisResults object can be used for more advanced
operations, such as exporting the results to a database or other formats
or use the native objects from other packages (e.g., patRoon) as we
demonstrate further in this article.
data.table objects
The features and feature groups can be obtained as
data.table with the
MassSpecEngine$get_features() and
MassSpecEngine$get_groups() methods. The methods also allow
to look for specific features/feature groups using mass, mass-to-charge
ratio, retention time and drift time targets, as show below for a small
set of compound targets where mass and retention time expected values
are known. Note that drift time is only applicable for MS data with ion
mobility separation.
db name formula mass rt tag
<char> <char> <num> <int> <char>
1: 4N-Acetylsulfadiazine C12H12N4O3S 292.0630 905 S
2: Metoprolol C15H25NO3 267.1834 915 S
3: Sulfamethoxazole C10H11N3O3S 253.0521 1015 S
4: Bisoprolol C18H31NO4 325.2253 955 S
5: 4N-Acetylsulfamethoxazole C12H13N3O4S 295.0627 1011 S
6: Carbamazepine C15H12N2O 236.0950 1079 S
7: Terbutryn C10H19N5S 241.1361 1126 S
8: Losartan C22H23ClN6O 422.1622 1095 S
9: Candesartan C24H20N6O3 440.1597 1097 S
10: Isoproturon C12H18N2O 206.1419 1152 S
11: Diuron C9H10Cl2N2O 232.0170 1160 S
12: Bezafibrat C19H20ClNO4 361.1081 1164 S
13: Valsartan C24H29N5O3 435.2270 1177 S
14: Tebuconazole C16H22ClN3O 307.1451 1267 S
15: Diclofenac C14H11Cl2NO2 295.0167 1255 S
16: Propiconazole C15H17Cl2N3O2 341.0698 1308 S
17: Flufenacet C14H13F4N3O2S 363.0665 1296 S
18: Ibuprofen C13H18O2 206.1307 1152 S
19: CBZD C15H14N2O3 270.1004 936 S
# Compounds are searched by monoisotopic mass and retention time
# ppm and sec set the mass (im ppm) and time (in seconds) allowed deviation, respectively
# average applies a mean to the intensities in each analysis replicate group
ms$get_groups(mass = db, ppm = 10, sec = 15, average = TRUE) group name effluent_neg effluent_pos influent_neg
<char> <char> <num> <num> <num>
1: M236_R1079_292 Carbamazepine 0.00 0.00 0.000
2: M253_R1015_479 Sulfamethoxazole 0.00 0.00 0.000
3: M267_R916_635 Metoprolol 0.00 15412.60 0.000
4: M295_R1256_1105 Diclofenac 0.00 0.00 8531.089
5: M325_R957_1538 Bisoprolol 0.00 13762.82 0.000
6: M440_R1097_2696 Candesartan 7415.23 22512.84 34920.953
influent_pos
<num>
1: 69645.159
2: 9809.779
3: 51331.574
4: 21566.104
5: 48375.419
6: 112540.651Already by inspection of the data.table, it is possible
to see compounds detected in the influent but not in the effluent (e.g.,
Carbamazepine) or compounds that are appear to be reduced during
ozonation (e.g., Metoprolol). Since positive and negative ionization
mode were combined, there are compounds that appear in both polarities
and are grouped by neutral monoisotopic mass (e.g., Candesartan and
Diclofenac).
plot_groups methods
For a better overview of the results, the method
MassSpecEngine$plot_groups() or even more detailed the
method MassSpecEngine$plot_groups_overview() can be
used.
# set legendNames to TRUE for using the names in db as legend
ms$plot_groups(mass = db, ppm = 10, sec = 15, legendNames = TRUE)
ms$plot_groups_overview(mass = db, ppm = 5, sec = 10, legendNames = TRUE)Filtered not removed
The FilterFeatures method was applied to filter features
according to defined conditions/thresholds. Yet, the filtered features
were not removed but just tagged as filtered. For instance, when the
method MassSpecEngine$get_groups() is run with
filtered argument set to TRUE, the filtered
features are also shown. Below, we search for the same compounds as
above but with the filtered argument set to
TRUE. Potential features from Valsartan are now returned
but were filtered due to low intensity. Note that when extracting
features based on basic parameters, i.e. mass and time, does not mean
that features are identified. The identification of features is a more
complex process and requires additional information, such as MS/MS data
as in the processing method suspect screening.
# Set filtered to TRUE for showing filtered features/feature groups
ms$get_groups(mass = db, ppm = 5, sec = 10, average = TRUE, filtered = TRUE) group name effluent_neg effluent_pos influent_neg
<char> <char> <num> <num> <num>
1: M236_R1079_292 Carbamazepine 0.000 0.00 0.000
2: M253_R1015_479 Sulfamethoxazole 0.000 0.00 0.000
3: M267_R916_635 Metoprolol 0.000 15412.60 0.000
4: M295_R1256_1105 Diclofenac 0.000 0.00 8531.089
5: M325_R957_1538 Bisoprolol 0.000 13762.82 0.000
6: M435_R1176_2663 Valsartan 3598.701 0.00 4133.250
7: M440_R1097_2696 Candesartan 7415.230 22512.84 34920.953
influent_pos
<num>
1: 69645.159
2: 9809.779
3: 51331.574
4: 21566.104
5: 48375.419
6: 0.000
7: 112540.651Internal Standards
The method FindInternalStandards was applied for tagging
spiked internal standards and the results can be obtained with the
dedicated method MassSpecEngine$get_internal_standards() or
plotted as a quality overview using the method
MassSpecEngine$plot_internal_standards(), as shown below.
The plot gives an overview of the mass, retention time and intensity
variance of the internal standards across the analyses in the
project.
# List of spiked internal standards
db_is name formula mass rt tag
<char> <char> <num> <int> <char>
1: Cyclophosphamide-D6 C7[2]H6H9Cl2N2O2P 266.0625 1007 IS
2: Ibuprofen-D3 C13[2]H3H15O2 209.1495 1150 IS
3: Diclophenac-D4 C14[2]H4H7Cl2NO2 299.0418 1253 IS
4: Metoprolol-D7 C15[2]H7H18NO3 274.2274 915 IS
5: Sulfamethoxazol-D4 C10[2]H4H7N3O3S 257.0772 1015 IS
6: Isoproturon-D6 C12[2]H6H12N2O 212.1796 1149 IS
7: Diuron-D6 C9[2]H6H4Cl2N2O 238.0547 1157 IS
8: Carbamazepin-D10 C15[2]H10H2N2O 246.1577 1075 IS
9: Naproxen-D3 C14[2]H3H11O3 233.1131 1169 IS
# Gets the internal standards evaluation data.table
ms$get_internal_standards()[, 1:7] name rt mass intensity area rtr mzr
<char> <num> <num> <num> <num> <num> <num>
1: Carbamazepin-D10 1074 246.1586 462051 2880482 14.4 0.0017
2: Carbamazepin-D10 1074 246.1585 539574 3525752 16.6 0.0021
3: Carbamazepin-D10 1075 246.1586 510457 3297997 15.5 0.0012
4: Cyclophosphamide-D6 1007 266.0627 51652 289270 13.5 0.0018
5: Cyclophosphamide-D6 1006 266.0627 56586 315716 10.2 0.0020
6: Cyclophosphamide-D6 1007 266.0628 60856 304027 9.4 0.0016
7: Diclophenac-D4 1254 299.0415 24569 149453 13.7 0.0018
8: Diclophenac-D4 1253 299.0423 52024 306126 12.5 0.0018
9: Diclophenac-D4 1255 299.0414 18342 120359 13.8 0.0027
10: Diclophenac-D4 1254 299.0424 48453 304167 13.4 0.0018
11: Diclophenac-D4 1254 299.0412 20070 127973 12.0 0.0027
12: Diclophenac-D4 1254 299.0423 51945 315610 12.0 0.0016
13: Diuron-D6 1157 238.0544 14884 87344 10.9 0.0010
14: Diuron-D6 1157 238.0553 120508 741890 12.7 0.0015
15: Diuron-D6 1157 238.0542 14522 85375 10.7 0.0018
16: Diuron-D6 1157 238.0553 130401 845728 13.4 0.0015
17: Diuron-D6 1157 238.0545 17604 102393 10.7 0.0017
18: Diuron-D6 1157 238.0554 141741 889026 14.8 0.0012
19: Isoproturon-D6 1149 212.1806 1198379 7589018 13.9 0.0012
20: Isoproturon-D6 1149 212.1808 1269561 8542166 16.4 0.0014
21: Isoproturon-D6 1149 212.1807 1322756 8587536 16.1 0.0016
22: Metoprolol-D7 915 274.2291 1581149 9737735 16.2 0.0017
23: Metoprolol-D7 914 274.2290 1497131 8861199 15.6 0.0026
24: Metoprolol-D7 915 274.2289 1571305 9472216 15.5 0.0013
25: Sulfamethoxazol-D4 1014 257.0766 10802 61534 9.6 0.0015
26: Sulfamethoxazol-D4 1014 257.0776 193814 1114649 14.5 0.0015
27: Sulfamethoxazol-D4 1014 257.0770 6566 29069 6.8 0.0020
28: Sulfamethoxazol-D4 1014 257.0777 189113 1099917 13.5 0.0015
29: Sulfamethoxazol-D4 1014 257.0770 7969 45062 10.0 0.0033
30: Sulfamethoxazol-D4 1014 257.0780 204834 1178733 14.1 0.0012
name rt mass intensity area rtr mzr
ms$plot_internal_standards()Quality control of spiked internal standards
ms$plot_groups_profile(mass = db_is, ppm = 8, sec = 10, filtered = TRUE, legendNames = TRUE)Internal standards profile across analyses
Components
The method AnnotateFeatures was applied to annotate the
natural isotopes and adducts into components. Implementation of
annotation for in-source fragments is planned but not yet available with
the StreamFind algorithm. The method
MassSpecEngine$get_components() can be used to search for
components, as shown below for the analysis number 11. Because the
filters excludeIsotopes and minIntensity were
applied, the isotopic features are likely filtered.
# Components of Diclofenac and Candesartan in analysis 11
components_example <- ms$get_components(
analyses = 11,
mass = db[db$name %in% c("Diclofenac", "Candesartan"), ],
ppm = 5, sec = 10,
filtered = TRUE
)
# Subset of the components data.table
components_example[, c(2:6, 8, 30:34)] replicate component feature group rt
<char> <char> <char> <char> <num>
1: influent_pos F388_MZ296_RT1256 F388_MZ296_RT1256 M295_R1256_1105 1256.670
2: influent_pos F388_MZ296_RT1256 F403_MZ298_RT1256 1255.600
3: influent_pos F937_MZ441_RT1097 F937_MZ441_RT1097 M440_R1097_2696 1096.581
4: influent_pos F937_MZ441_RT1097 F940_MZ442_RT1097 1096.580
5: influent_pos F937_MZ441_RT1097 F946_MZ443_RT1097 M442_R1098_2714 1096.581
intensity i.component_feature iso_size iso_charge iso_step iso_cat
<num> <char> <int> <int> <int> <char>
1: 19179.480 F388_MZ296_RT1256 2 1 0 M+0
2: 14199.290 F388_MZ296_RT1256 2 1 2 M+2
3: 109887.297 F937_MZ441_RT1097 3 1 0 M+0
4: 28418.029 F937_MZ441_RT1097 3 1 1 M+1
5: 4544.722 F937_MZ441_RT1097 3 1 2 M+2The components (i.e., isotopes and adducts) can also be visualized
with the method MassSpecEngine$map_components(), as shown
below for the internal standards added to analysis 11. Note that again
the filtered argument was set to TRUE to
return also filtered features, as the internal standards were likely
excluded by blank subtraction.
ms$map_components(
analyses = 11,
mass = db_is,
ppm = 8, sec = 10,
filtered = TRUE,
legendNames = TRUE
)Components (i.e., isotopes and adducts) of internal standards in analysis 11
Suspects
The methods MassSpecEngine$get_suspects() and
MassSpecEngine$plot_suspects() can be used to inspect the
suspect screening results. In the plot function, a second plot is added
to compare the experimental fragmentation pattern (top) with the
fragmentation pattern of the respective reference standard (down) added
within the database. The colorBy argument can be set to
targets+replicates to legend the plot with combined keys of
suspect target names and analysis replicate names.
ms$get_suspects()[, c(1, 5, 14)] analysis name id_level
<char> <char> <char>
1: 02_tof_ww_is_pos_influent-r001 Carbamazepine 1
2: 02_tof_ww_is_pos_influent-r001 Sulfamethoxazole 1
3: 02_tof_ww_is_pos_influent-r001 Metoprolol 1
4: 02_tof_ww_is_pos_influent-r001 Diclofenac 1
5: 02_tof_ww_is_pos_influent-r001 Bisoprolol 1
6: 02_tof_ww_is_pos_influent-r001 Candesartan 1
7: 02_tof_ww_is_pos_influent-r002 Carbamazepine 1
8: 02_tof_ww_is_pos_influent-r002 Sulfamethoxazole 3b
9: 02_tof_ww_is_pos_influent-r002 Metoprolol 1
10: 02_tof_ww_is_pos_influent-r002 Diclofenac 1
11: 02_tof_ww_is_pos_influent-r002 Bisoprolol 1
12: 02_tof_ww_is_pos_influent-r002 Candesartan 1
13: 02_tof_ww_is_pos_influent-r003 Carbamazepine 1
14: 02_tof_ww_is_pos_influent-r003 Sulfamethoxazole 1
15: 02_tof_ww_is_pos_influent-r003 Metoprolol 1
16: 02_tof_ww_is_pos_influent-r003 Diclofenac 3b
17: 02_tof_ww_is_pos_influent-r003 Bisoprolol 1
18: 02_tof_ww_is_pos_influent-r003 Candesartan 1
19: 03_tof_ww_is_pos_o3sw_effluent-r001 Metoprolol 1
20: 03_tof_ww_is_pos_o3sw_effluent-r001 Bisoprolol 1
21: 03_tof_ww_is_pos_o3sw_effluent-r001 Candesartan 1
22: 03_tof_ww_is_pos_o3sw_effluent-r002 Metoprolol 1
23: 03_tof_ww_is_pos_o3sw_effluent-r002 Bisoprolol 1
24: 03_tof_ww_is_pos_o3sw_effluent-r002 Candesartan 1
25: 03_tof_ww_is_pos_o3sw_effluent-r003 Metoprolol 1
26: 03_tof_ww_is_pos_o3sw_effluent-r003 Bisoprolol 1
27: 03_tof_ww_is_pos_o3sw_effluent-r003 Candesartan 1
analysis name id_level
ms$plot_suspects(colorBy = "targets+replicates")Methods from patRoon
The NonTargetAnalysisResults object holds methods to obtain original
objects from the patRoon package. For
instance, the S4 class features or
featureGroups objects can be obtained via the
get_patRoon_features method of the
NonTargetAnalysisResults results. The patRoon
package provides a comprehensive set of functions, as shown below. See
more information in the patRoon
reference documentation.
# Native patRoon object
fGroups <- get_patRoon_features(ms$NonTargetAnalysisResults, filtered = FALSE, featureGroups = TRUE)
fGroupsA featureGroupsSet object
Hierarchy:
featureGroups
|-- featureGroupsSet
|-- featureGroupsScreeningSet
---
Object size (indication): 26.5 MB
Algorithm: openms-set
Feature groups: M236_R1236_301, M236_R1222_300, M236_R1139_302, M240_R945_325, M240_R966_329, M242_R916_347, ... (374 total)
Features: 1662 (4.4 per group)
Has normalized intensities: FALSE
Internal standards used for normalization: no
Predicted concentrations: none
Predicted toxicities: none
Analyses: 02_tof_ww_is_neg_influent-r001, 02_tof_ww_is_neg_influent-r002, 02_tof_ww_is_neg_influent-r003, 03_tof_ww_is_neg_o3sw_effluent-r001, 03_tof_ww_is_neg_o3sw_effluent-r002, 03_tof_ww_is_neg_o3sw_effluent-r003, ... (12 total)
Replicate groups: influent_neg, effluent_neg, influent_pos, effluent_pos (4 total)
Replicate groups used as blank: blank_neg, blank_pos (2 total)
Sets: negative, positive
# Using the native patRoon's plotUpSet method
patRoon::plotUpSet(fGroups)
Fold-change analysis
The method get_fold_change() and correspondent
plot_fold_change() can be used to calculate and plot the
fold-change between influent and effluent samples. The method calculates
the fold-change for each feature group and replicates, as shown below.
The plot shows the fold-change for each feature group across the
replicates. The fold-change is calculated according to Bader et
al. (2017), leveraging the replicates variance to increase the
significance of the fold-change. Formation is not in the plot as no new
features were detected in the effluent of the wastewater ozonation
treatment step.
ms$plot_fold_change(
replicatesIn = c("influent_neg", "influent_pos"),
replicatesOut = c("effluent_neg", "effluent_pos"),
filtered = FALSE,
constantThreshold = 0.5,
eliminationThreshold = 0.2,
correctIntensity = TRUE,
fillZerosWithLowerLimit = FALSE, # set to TRUE for filling zeros with lower limit argument
lowerLimit = 200
)Alternatively, the plotVolcano from patRoon can be used to
plot fold-change. For more information check the patRoon
reference documentation. Note that the function
patRoon::getFCParams is used to define the parameters for
the fold-change calculation. Below we use the default argument values,
only the in and out replicate names are set.
patRoon::plotVolcano(
fGroups, # obtained above with get_patRoon_features
patRoon::getFCParams(c("influent_pos", "effluent_pos"))
)