... | ... |
@@ -34,21 +34,16 @@ Predicts new paths given a pathCluster model. |
34 | 34 |
weight.method = "cor", use.attr="miriam.uniprot", bootstrap = FALSE) |
35 | 35 |
|
36 | 36 |
## Get ranked paths using probabilistic shortest paths. |
37 |
- ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
|
37 |
+ ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
|
38 | 38 |
K=20, minPathSize=8) |
39 |
- |
|
40 |
- ## Convert paths to binary matrix. |
|
39 |
+ |
|
40 |
+ ## Convert paths to binary matrix. |
|
41 | 41 |
ybinpaths <- pathsToBinary(ranked.p) |
42 | 42 |
p.cluster <- pathCluster(ybinpaths, M=2) |
43 | 43 |
|
44 | 44 |
## just an example of how to predict cluster membership. |
45 | 45 |
pclust.pred <- predictPathCluster(p.cluster,ybinpaths$paths) |
46 |
- |
|
47 |
-} |
|
48 |
-\author{ |
|
49 |
-Ichigaku Takigawa |
|
50 | 46 |
|
51 |
-Timothy Hancock |
|
52 | 47 |
} |
53 | 48 |
\seealso{ |
54 | 49 |
Other Path clustering & classification methods: \code{\link{pathClassifier}}, |
... | ... |
@@ -59,4 +54,9 @@ Other Path clustering & classification methods: \code{\link{pathClassifier}}, |
59 | 54 |
\code{\link{plotPathCluster}}, |
60 | 55 |
\code{\link{predictPathClassifier}} |
61 | 56 |
} |
57 |
+\author{ |
|
58 |
+Ichigaku Takigawa |
|
62 | 59 |
|
60 |
+Timothy Hancock |
|
61 |
+} |
|
62 |
+\concept{Path clustering & classification methods} |
... | ... |
@@ -1,4 +1,5 @@ |
1 |
-% Generated by roxygen2 (4.0.1): do not edit by hand |
|
1 |
+% Generated by roxygen2: do not edit by hand |
|
2 |
+% Please edit documentation in R/pathCluster.R |
|
2 | 3 |
\name{predictPathCluster} |
3 | 4 |
\alias{predictPathCluster} |
4 | 5 |
\title{Predicts new paths given a pathCluster model} |
... | ... |
@@ -21,7 +22,7 @@ A list with the following elements: |
21 | 22 |
Predicts new paths given a pathCluster model. |
22 | 23 |
} |
23 | 24 |
\examples{ |
24 |
-## Prepare a weighted reaction network. |
|
25 |
+ ## Prepare a weighted reaction network. |
|
25 | 26 |
## Conver a metabolic network to a reaction network. |
26 | 27 |
data(ex_sbml) # bipartite metabolic network of Carbohydrate metabolism. |
27 | 28 |
rgraph <- makeReactionNetwork(ex_sbml, simplify=TRUE) |
... | ... |
@@ -33,15 +34,16 @@ Predicts new paths given a pathCluster model. |
33 | 34 |
weight.method = "cor", use.attr="miriam.uniprot", bootstrap = FALSE) |
34 | 35 |
|
35 | 36 |
## Get ranked paths using probabilistic shortest paths. |
36 |
- ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
|
37 |
+ ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
|
37 | 38 |
K=20, minPathSize=8) |
38 |
- |
|
39 |
- ## Convert paths to binary matrix. |
|
39 |
+ |
|
40 |
+ ## Convert paths to binary matrix. |
|
40 | 41 |
ybinpaths <- pathsToBinary(ranked.p) |
41 | 42 |
p.cluster <- pathCluster(ybinpaths, M=2) |
42 | 43 |
|
43 | 44 |
## just an example of how to predict cluster membership. |
44 | 45 |
pclust.pred <- predictPathCluster(p.cluster,ybinpaths$paths) |
46 |
+ |
|
45 | 47 |
} |
46 | 48 |
\author{ |
47 | 49 |
Ichigaku Takigawa |
... | ... |
@@ -49,14 +51,12 @@ Ichigaku Takigawa |
49 | 51 |
Timothy Hancock |
50 | 52 |
} |
51 | 53 |
\seealso{ |
52 |
-Other Path clustering & classification methods: \code{\link{pathClassifier}}; |
|
53 |
- \code{\link{pathCluster}}; \code{\link{pathsToBinary}}; |
|
54 |
- \code{\link{plotClassifierROC}}; |
|
54 |
+Other Path clustering & classification methods: \code{\link{pathClassifier}}, |
|
55 |
+ \code{\link{pathCluster}}, \code{\link{pathsToBinary}}, |
|
56 |
+ \code{\link{plotClassifierROC}}, |
|
55 | 57 |
\code{\link{plotClusterMatrix}}, |
56 |
- \code{\link{plotClusterProbs}}, |
|
57 |
- \code{\link{plotClusters}}; |
|
58 |
- \code{\link{plotPathClassifier}}; |
|
59 |
- \code{\link{plotPathCluster}}; |
|
58 |
+ \code{\link{plotPathClassifier}}, |
|
59 |
+ \code{\link{plotPathCluster}}, |
|
60 | 60 |
\code{\link{predictPathClassifier}} |
61 | 61 |
} |
62 | 62 |
|
added plotCytoscapeGML , Geneset export formats, bug fixes
... | ... |
@@ -1,3 +1,4 @@ |
1 |
+% Generated by roxygen2 (4.0.1): do not edit by hand |
|
1 | 2 |
\name{predictPathCluster} |
2 | 3 |
\alias{predictPathCluster} |
3 | 4 |
\title{Predicts new paths given a pathCluster model} |
... | ... |
@@ -5,19 +6,16 @@ |
5 | 6 |
predictPathCluster(pfit, newdata) |
6 | 7 |
} |
7 | 8 |
\arguments{ |
8 |
- \item{pfit}{The pathway cluster model trained by |
|
9 |
- \code{\link{pathCluster}} or |
|
10 |
- \code{\link{pathClassifier}}.} |
|
9 |
+\item{pfit}{The pathway cluster model trained by \code{\link{pathCluster}} or \code{\link{pathClassifier}}.} |
|
11 | 10 |
|
12 |
- \item{newdata}{The binary pathway dataset to be assigned |
|
13 |
- a cluster label.} |
|
11 |
+\item{newdata}{The binary pathway dataset to be assigned a cluster label.} |
|
14 | 12 |
} |
15 | 13 |
\value{ |
16 |
-A list with the following elements: \tabular{ll}{ |
|
17 |
-\code{labels} \tab a vector indicating the 3M cluster |
|
18 |
-membership. \cr \code{posterior.probs} \tab a matrix of |
|
19 |
-posterior probabilities for each path belonging to each |
|
20 |
-cluster. } |
|
14 |
+A list with the following elements: |
|
15 |
+\tabular{ll}{ |
|
16 |
+\code{labels} \tab a vector indicating the 3M cluster membership. \cr |
|
17 |
+\code{posterior.probs} \tab a matrix of posterior probabilities for each path belonging to each cluster. |
|
18 |
+} |
|
21 | 19 |
} |
22 | 20 |
\description{ |
23 | 21 |
Predicts new paths given a pathCluster model. |
... | ... |
@@ -51,14 +49,14 @@ Ichigaku Takigawa |
51 | 49 |
Timothy Hancock |
52 | 50 |
} |
53 | 51 |
\seealso{ |
54 |
-Other Path clustering & classification methods: |
|
55 |
-\code{\link{pathClassifier}}; \code{\link{pathCluster}}; |
|
56 |
-\code{\link{pathsToBinary}}; |
|
57 |
-\code{\link{plotClassifierROC}}; |
|
58 |
-\code{\link{plotClusterMatrix}}, |
|
59 |
-\code{\link{plotClusterProbs}}, \code{\link{plotClusters}}; |
|
60 |
-\code{\link{plotPathClassifier}}; |
|
61 |
-\code{\link{plotPathCluster}}; |
|
62 |
-\code{\link{predictPathClassifier}} |
|
52 |
+Other Path clustering & classification methods: \code{\link{pathClassifier}}; |
|
53 |
+ \code{\link{pathCluster}}; \code{\link{pathsToBinary}}; |
|
54 |
+ \code{\link{plotClassifierROC}}; |
|
55 |
+ \code{\link{plotClusterMatrix}}, |
|
56 |
+ \code{\link{plotClusterProbs}}, |
|
57 |
+ \code{\link{plotClusters}}; |
|
58 |
+ \code{\link{plotPathClassifier}}; |
|
59 |
+ \code{\link{plotPathCluster}}; |
|
60 |
+ \code{\link{predictPathClassifier}} |
|
63 | 61 |
} |
64 | 62 |
|
... | ... |
@@ -36,11 +36,11 @@ Predicts new paths given a pathCluster model. |
36 | 36 |
|
37 | 37 |
## Get ranked paths using probabilistic shortest paths. |
38 | 38 |
ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
39 |
- K=20, minPathSize=6) |
|
39 |
+ K=20, minPathSize=8) |
|
40 | 40 |
|
41 | 41 |
## Convert paths to binary matrix. |
42 | 42 |
ybinpaths <- pathsToBinary(ranked.p) |
43 |
- p.cluster <- pathCluster(ybinpaths, M=3) |
|
43 |
+ p.cluster <- pathCluster(ybinpaths, M=2) |
|
44 | 44 |
|
45 | 45 |
## just an example of how to predict cluster membership. |
46 | 46 |
pclust.pred <- predictPathCluster(p.cluster,ybinpaths$paths) |
1 | 1 |
new file mode 100644 |
... | ... |
@@ -0,0 +1,64 @@ |
1 |
+\name{predictPathCluster} |
|
2 |
+\alias{predictPathCluster} |
|
3 |
+\title{Predicts new paths given a pathCluster model} |
|
4 |
+\usage{ |
|
5 |
+predictPathCluster(pfit, newdata) |
|
6 |
+} |
|
7 |
+\arguments{ |
|
8 |
+ \item{pfit}{The pathway cluster model trained by |
|
9 |
+ \code{\link{pathCluster}} or |
|
10 |
+ \code{\link{pathClassifier}}.} |
|
11 |
+ |
|
12 |
+ \item{newdata}{The binary pathway dataset to be assigned |
|
13 |
+ a cluster label.} |
|
14 |
+} |
|
15 |
+\value{ |
|
16 |
+A list with the following elements: \tabular{ll}{ |
|
17 |
+\code{labels} \tab a vector indicating the 3M cluster |
|
18 |
+membership. \cr \code{posterior.probs} \tab a matrix of |
|
19 |
+posterior probabilities for each path belonging to each |
|
20 |
+cluster. } |
|
21 |
+} |
|
22 |
+\description{ |
|
23 |
+Predicts new paths given a pathCluster model. |
|
24 |
+} |
|
25 |
+\examples{ |
|
26 |
+## Prepare a weighted reaction network. |
|
27 |
+ ## Conver a metabolic network to a reaction network. |
|
28 |
+ data(ex_sbml) # bipartite metabolic network of Carbohydrate metabolism. |
|
29 |
+ rgraph <- makeReactionNetwork(ex_sbml, simplify=TRUE) |
|
30 |
+ |
|
31 |
+ ## Assign edge weights based on Affymetrix attributes and microarray dataset. |
|
32 |
+ # Calculate Pearson's correlation. |
|
33 |
+ data(ex_microarray) # Part of ALL dataset. |
|
34 |
+ rgraph <- assignEdgeWeights(microarray = ex_microarray, graph = rgraph, |
|
35 |
+ weight.method = "cor", use.attr="miriam.uniprot", bootstrap = FALSE) |
|
36 |
+ |
|
37 |
+ ## Get ranked paths using probabilistic shortest paths. |
|
38 |
+ ranked.p <- pathRanker(rgraph, method="prob.shortest.path", |
|
39 |
+ K=20, minPathSize=6) |
|
40 |
+ |
|
41 |
+ ## Convert paths to binary matrix. |
|
42 |
+ ybinpaths <- pathsToBinary(ranked.p) |
|
43 |
+ p.cluster <- pathCluster(ybinpaths, M=3) |
|
44 |
+ |
|
45 |
+ ## just an example of how to predict cluster membership. |
|
46 |
+ pclust.pred <- predictPathCluster(p.cluster,ybinpaths$paths) |
|
47 |
+} |
|
48 |
+\author{ |
|
49 |
+Ichigaku Takigawa |
|
50 |
+ |
|
51 |
+Timothy Hancock |
|
52 |
+} |
|
53 |
+\seealso{ |
|
54 |
+Other Path clustering & classification methods: |
|
55 |
+\code{\link{pathClassifier}}; \code{\link{pathCluster}}; |
|
56 |
+\code{\link{pathsToBinary}}; |
|
57 |
+\code{\link{plotClassifierROC}}; |
|
58 |
+\code{\link{plotClusterMatrix}}, |
|
59 |
+\code{\link{plotClusterProbs}}, \code{\link{plotClusters}}; |
|
60 |
+\code{\link{plotPathClassifier}}; |
|
61 |
+\code{\link{plotPathCluster}}; |
|
62 |
+\code{\link{predictPathClassifier}} |
|
63 |
+} |
|
64 |
+ |