| Version: | 0.5 |
| Date: | 2025-06-04 |
| Maintainer: | Stefano Cacciatore <tkcaccia@gmail.com> |
| Title: | Chemical Structural Similarity Analysis |
| Description: | A new pipeline to explore chemical structural similarity across metabolites. It allows the metabolite classification in structurally-related modules and identifies common shared functional groups. The KODAMA algorithm is used to highlight structural similarity between metabolites. See Cacciatore S, Tenori L, Luchinat C, Bennett PR, MacIntyre DA. (2017) Bioinformatics <doi:10.1093/bioinformatics/btw705>, Cacciatore S, Luchinat C, Tenori L. (2014) Proc Natl Acad Sci USA <doi:10.1073/pnas.1220873111>, and Abdel-Shafy EA, Melak T, MacIntyre DA, Zadra G, Zerbini LF, Piazza S, Cacciatore S. (2023) Bioinformatics Advances <doi:10.1093/bioadv/vbad053>. |
| Depends: | R (≥ 3.5.0), stats, KODAMA (≥ 3.0), httr, XML, fingerprint, rcdk (≥ 3.4.3) |
| Suggests: | knitr, rmarkdown, readxl, impute, pheatmap, RColorBrewer, clinical |
| VignetteBuilder: | knitr |
| SuggestsNote: | No suggestions |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| Packaged: | 2025-06-04 15:53:53 UTC; user |
| NeedsCompilation: | no |
| Repository: | CRAN |
| Author: | Ebtesam Abdel-Shafy [aut], Tadele Melak [aut], David A. MacIntyre [aut], Giorgia Zadra [aut], Luiz F. Zerbini [aut], Silvano Piazza [aut], Stefano Cacciatore [aut, cre] |
| Date/Publication: | 2025-06-04 16:30:02 UTC |
ChemRICH Dataset
Description
This dataset consists of a list of the metabolites names download from https://chemrich.fiehnlab.ucdavis.edu/. HMDB IDs were retrieved from PubChem Identifier Exchange Service (https://pubchem.ncbi.nlm.nih.gov/idexchange/idexchange.cgi) and manually curated.
Usage
data(ChemRICH)Value
A list with the following elements in the variable ChemRICH:
name |
A vector of metabolite's names. |
SMILES |
A vector of SMILES represenation of each metabolite. |
HMDB |
A vector containing HMDB IDs of each metabolite. |
Examples
data(ChemRICH)
HFD Dataset
Description
This dataset is dataframe of metabolite dataset contains only chemical information.
Usage
data(HFD)Value
A list with the following elements in the variable HFD:
SMILES |
A vector of SMILES represenation of each metabolite. |
CHEMICAL_ID |
A vector of chemical ID number or each metabolite. |
PUBCHEM |
A vector of identifier ID number from PUBCHEM database for chemical molecules and their activities in biological assays. |
CHEMSPIDER |
A vector of a unique identifier from CHEMSPIDER database each molecule. |
HMDB |
A vector containing HMDB IDs of each metabolite. |
Examples
data(HFD)
KODAMA chemical similarity.
Description
This function calculates the structural similarity between different metabolites and performs hierarchical clustering using the KODAMA algorithm.
Usage
KODAMA.chem.sim (smiles,
d=NULL,
k=50,
dissimilarity.parameters=list(),
kodama.matrix.parameters=list(),
kodama.visualization.parameters=list(),
hclust.parameters=list(method="ward.D"))
Arguments
smiles |
A list of smile notations for the study metabolites dataset. |
d |
A distance structure such as that returned by dist or a full symmetric matrix containing the dissimilarities. If NULL (default), then the dissimilarity matrix will be generated by |
k |
A number of components of multidimensional scaling. |
dissimilarity.parameters |
Optional parameters for |
kodama.matrix.parameters |
Optional parameters for |
kodama.visualization.parameters |
Optional parameters for |
hclust.parameters |
Optional parameters for |
Value
A list contains all results of KODAMA chemical similarity analysis and hierarchical clustering for the KODAMA dimensions.
See Also
KODAMA.matrix, KODAMA.visualization,
Examples
data(Metabolites)
res=KODAMA.chem.sim(Metabolites$SMILES)
plot(res$kodama$visualization)
Metabolomic Dataset
Description
This dataset consists of a list of the metabolites as returned by the function readMet and concentration value of each metabolites.
Usage
data(Metabolites)Value
A list with the following elements in the variable Metabolites:
concentration |
A matrix containing the concentration of each metabolites. |
name |
A vector of metabolite's names. |
SMILES |
A vector of SMILES represenation of each metabolite. |
HMDB |
A vector containing HMDB IDs of each metabolite. |
readMet |
A list of metabolites information produced by |
Examples
data(Metabolites)
Weighted Metabolite Chemical Structural Analysis
Description
Summarize metabolites concetration in each of identified clusters using the module eigenvalue (eigen-metabolite) for calculating module membership measures.
Usage
WMCSA(data,cl)
Arguments
data |
dataset of different metabolite concentration in differnt samples. |
cl |
The output of the |
Value
This function returns a matrix as output represent similarity score of metabolites within the same module among different samples.
See Also
Examples
data(Metabolites)
SMILES=Metabolites$SMILES
names(SMILES)=Metabolites$name
res=KODAMA.chem.sim(SMILES)
cl=allbranches(res$hclust)
ww=WMCSA(Metabolites$concentration,cl)
Cut a Tree into Groups of Data
Description
Cuts a tree as resulting from hclust function, into groups (a.k.a. modules).
Usage
allbranches(hh,minlen=5)
Arguments
hh |
a tree as produced by |
minlen |
The minimum number of elements in each module. |
Value
A list contains vectors of module memberships.
See Also
cutree, hclust, clusters.detection
Examples
data(Metabolites)
data=Metabolites$readMet$concentration
hh=hclust(dist(data),method="ward.D")
res=allbranches(hh)
Chemical dissimilarity.
Description
This function calculates the structural dissimilarity between different metabolites using the simplified molecular-input line-entry system (SMILE) of each metabolite as input.
Usage
chemical.dissimilarity (smiles,method="tanimoto",type="extended")
Arguments
smiles |
A vector of smile notations. |
method |
The method used to calculated the distance between molecular fingerprint ("tanimoto" as default). For more information see |
type |
The type of fingerprint applied to the SMILEs ("extended" as default). For more information see |
Value
A list contains distance between fingerprints .
See Also
fp.sim.matrix, get.fingerprint,
Examples
data(Metabolites)
d=chemical.dissimilarity(Metabolites$SMILES[1:50])
Detection of clusters.
Description
This function calculates the structural similarity between different metabolites, performs hierarchical clustering using the KODAMA algorithm, and detects the optimal number of clusters. The procedure is repeated to ensure the robustness of the detection.
Usage
clusters.detection (smiles,
k=50,
seed=12345,
max_nc = 30,
dissimilarity.parameters=list(),
kodama.matrix.parameters=list(),
kodama.visualization.parameters=list(),
hclust.parameters=list(method="ward.D"),
verbose = TRUE)
Arguments
smiles |
A list of smile notations for the study metabolites dataset. |
k |
The number of components of multidimensional scaling. |
seed |
Seed for the generation of random numbers. |
max_nc |
Maximum number of clusters. |
dissimilarity.parameters |
Optional parameters for |
kodama.matrix.parameters |
Optional parameters for |
kodama.visualization.parameters |
Optional parameters for |
hclust.parameters |
Optional parameters for |
verbose |
If verbose is TRUE, it displays the progress for each iteration. |
Value
A list contains all results of KODAMA chemical similarity analysis and hierarchical clustering.
See Also
KODAMA.matrix, KODAMA.visualization
Examples
data(Metabolites)
res=clusters.detection(Metabolites$SMILES)
Metabolite-associated Diseases
Description
This function correlates metabolites to associated diseases.
Usage
diseasesMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the diseases associated with each metabolite.
See Also
pathwaysMet, taxonomyMet, enzymesMet
Examples
data(Metabolites)
dis=diseasesMet(Metabolites$readMet)
Metabolite-associated Enzymes
Description
This function finds the metabolite related enzymes.
Usage
enzymesMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the enzymes associated with each metabolite.
See Also
pathwaysMet , taxonomyMet, diseasesMet
Examples
data(Metabolites)
enz=enzymesMet(Metabolites$readMet)
Cluster features extraction
Description
This function finds features associated with each cluster.
Usage
features(doc,cla,cl,HMDB_ID)
Arguments
doc |
The output of the |
cla |
The output of |
cl |
The output of the |
HMDB_ID |
A vector of HMDB IDs associated with their chemical name. |
Value
A list of p-value calculated using Fisher test for cluster associted features.
See Also
KODAMA.chem.sim, tree.cutting, substituentsMet
Examples
data(Metabolites)
SMILES=Metabolites$SMILES
names(SMILES)=Metabolites$name
HMDB=Metabolites$HMDB
names(HMDB)=Metabolites$name
res=KODAMA.chem.sim(SMILES)
cl=allbranches(res$hclust)
cla=substituentsMet(Metabolites$readMet)
f=features(Metabolites$readMet,cla,cl,HMDB)
Name of metabolites
Description
This function extracts the metabolite's names from the list generated by readMet function.
Usage
nameMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the names of each metabolite.
See Also
Examples
data(Metabolites)
nam=nameMet(Metabolites$readMet)
Metabolic Pathways
Description
This function finds the metabolite related pathways.
Usage
pathwaysMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the pathways associated with each metabolite.
See Also
readMet, taxonomyMet, enzymesMet, diseasesMet
Examples
data(Metabolites)
pat=pathwaysMet(Metabolites$readMet)
Physical Proprieties of metabolites
Description
This function finds the Physical Proprieties of metabolites.
Usage
propertiesMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the properties associated with each metabolite.
See Also
readMet, taxonomyMet, substituentsMet, propertiesMet
Examples
data(Metabolites)
pro=propertiesMet(Metabolites$readMet)
Metabolite Cards Reading
Description
This function extract metabocards of your metabolites dataset from http://www.hmdb.ca/metabolites/ database and store all of this information in a list.
Usage
readMet(ID, address = c("http://www.hmdb.ca/metabolites/"),remove=TRUE)
Arguments
ID |
A vector containg the HMDBcodes (i.e., metabolite IDs) of metabolites dataset. |
address |
Optional address where the MetaboCards are located. The default address is http://www.hmdb.ca/metabolites/. |
remove |
A logic value. If true, missing and wrong HMDB IDs are removed. |
Value
A list containing all the information related to the metabocards.
See Also
Examples
ID=c("HMDB0000122","HMDB0000124","HMDB0000243","HMDB0000263")
doc=readMet(ID)
Metabolites selection
Description
This function select metabolites from the list generated by readMet function.
Usage
selectionMet(doc, sel)
Arguments
doc |
A list of metabolites information produced by |
sel |
A vector of metabolite's HMDBcode that will be selected |
Value
doc |
A doc list contains only the selcted metabolites. |
See Also
Examples
data(Metabolites)
doc=selectionMet(Metabolites$readMet,c("HMDB0000299","HMDB0000881"))
nameMet(doc)
Metabolite substituents
Description
This function finds the metabolite related substituents.
Usage
substituentsMet(doc)
Arguments
doc |
A list of metabolites information produced by |
.
Value
A data frame contains the substituents of each metabolite.
See Also
readMet, nameMet, propertiesMet
Examples
data(Metabolites)
sub=substituentsMet(Metabolites$readMet)
Metabolite Taxonomy
Description
This function finds the metabolite related taxonomy.
Usage
taxonomyMet(doc)
Arguments
doc |
A list of metabolites information produced by |
Value
A data frame contains the taxonomy of each metabolite.
See Also
readMet, propertiesMet, enzymesMet, diseasesMet
Examples
data(Metabolites)
tax=taxonomyMet(Metabolites$readMet)
Optimal cluster number calculation.
Description
This function helps to estimate the optimal cluster number that fit the metabolites dataset. It applies different optimal cluster number calculating algorithms to cut clutering tree of hclust function. and return a list contains index corresponding to each cluster number.
Usage
tree.cutting (res,max_nc=20)
Arguments
res |
A list produced by |
max_nc |
The maximum number of cluster (default = 20). |
Value
A list contains the calculation for each clustering of Rousseeuw's Silhouette index.
See Also
Examples
data(Metabolites)
res=KODAMA.chem.sim(Metabolites$SMILES)
clu=tree.cutting(res,max_nc = 30)
plot(clu$min_nc:clu$max_nc,clu$res.S)
Write a CLS file
Description
This function write a file in the format CLS defined by GenePattern.
Usage
write.cls(es, address)
Arguments
es |
A matrix. |
address |
The address of the file should be saved. |
Value
No return value. If an invalid address is inserted, the function will generate an error.
See Also
Write a GCT file
Description
This function write a file in the format GCT defined by GenePattern.
Usage
write.gct(es, address)
Arguments
es |
A matrix. |
address |
The address of the file should be saved. |
Value
No return value. If an invalid address is inserted, the function will generate an error.
See Also
Write a GMT file
Description
This function write a file containing the Metabolite Set informtation in the format GMT defined by GenePattern.
Usage
write.gmt(sub,address,min_entry=2,max_entry=50)
Arguments
sub |
A matrix. |
address |
The address of the file should be saved. |
min_entry |
The minimum number of metabolites for each metabolite set. |
max_entry |
The maximum number of metabolites for each metabolite set. |
Value
No return value. If an invalid address is inserted, the function will generate an error.