1. TCGAbiolinks
  2. Commit 4162fa4afaea5e3602865f581ec70bd64d0b54eb
Browse code

Add glioma classifier function from the GUI

Former-commit-id: 87ca26c2223fc6bc55154728dfbd6e5870323992 [formerly 986f30d2ccb71876b4062b03eb705d5523a01ea6]
Former-commit-id: a29c5eba559c2f955f206204b8e756546bfaaedb

Tiago Chedraoui Silva authored on 27/11/2019 18:14:10
Showing 3 changed files
tests/testthat/test-analyse.R
History View file @ 4162fa4a
... ... 
@@ -83,7 +83,7 @@ test_that("TCGAanalyze_DMC is handling NAs correctly", {
83 83 
         c(rep("group1", 10),  rep("group2", 10))
84 84 
     hypo.hyper <-
85 85 
         TCGAanalyze_DMC(data, p.cut = 0.85, "group", "group1", "group2")
86 
-    result <- hypo.hyper
86 
+    result <- hypo.hyper[1,]
87 87 
     expect_equal(result$mean.group1, 0.9)
88 88 
     expect_equal(result$mean.group2, 0.1)
89 89 
     expect_equal(result$mean.group1.minus.mean.group2 , 0.8)
vignettes/_site.yml
History View file @ 4162fa4a
... ... 
@@ -32,13 +32,21 @@ navbar:
32 32 
           icon: fa-flask
33 33 
           href: http://rpubs.com/tiagochst/atac_seq_workshop
34 34 
     - text: "Analysis"
35 
-      href: analysis.html
36 35 
       icon: fa-flask
36 
+      menu:
37 
+        - text: "---------"
38 
+        - text: "Analysis functions"
39 
+          icon: fa-flask
40 
+          href: analysis.html
41 
+        - text: "Other functions"
42 
+          icon: fa-flask
43 
+          href: extension.html
44 
+        - text: "---------"
45 
+        - text: "Glioma classsifier"
46 
+          icon: fa-bulleye
47 
+          href: classifiers.html
37 48 
     - text: "Case Study"
38 49 
       href: casestudy.html
39 
-    - text: "Other functions"
40 
-      icon: fa-book
41 
-      href: extension.html
42 50 
     - text: "GUI"
43 51 
       icon: fa-hand-pointer-o
44 52 
       href: gui.html
vignettes/classifiers.Rmd
History View file @ 4162fa4a
45 53 
new file mode 100755
... ... 
@@ -0,0 +1,93 @@
1 
+---
2 
+title: "Classifiers methods"
3 
+bibliography: bibliography.bib
4 
+vignette: >
5 
+    %\VignetteIndexEntry{10. Classifiers}
6 
+    %\VignetteEngine{knitr::rmarkdown}
7 
+---
8 
+
9 
+```{r setup, include=FALSE}
10 
+knitr::opts_chunk$set(dpi = 300)
11 
+knitr::opts_chunk$set(cache = FALSE)
12 
+```
13 
+
14 
+```{r, echo = FALSE,hide=TRUE, message=FALSE,warning=FALSE}
15 
+devtools::load_all(".")
16 
+```
17 
+
18 
+```{r message=FALSE, warning=FALSE, include=FALSE}
19 
+library(SummarizedExperiment)
20 
+library(dplyr)
21 
+library(DT)
22 
+```
23 
+
24 
+<br>
25 
+
26 
+## Classifying gliomas samples with `gliomaClassifier`
27 
+
28 
+Classify glioma samples with DNA methylation array based on:
29 
+
30 
+**Ceccarelli, Michele, et al. "Molecular profiling reveals biologically discrete subsets and pathways of progression in diffuse glioma." Cell 164.3 (2016): 550-563.** (https://doi.org/10.1016/j.cell.2015.12.028)
31 
+
32 
+Possible classifications are:
33 
+
34 
+- Mesenchymal-like
35 
+- Classic-like
36 
+- G-CIMP-high
37 
+- G-CIMP-low
38 
+- LGm6-GBM
39 
+- Codel
40 
+
41 
+### Data
42 
+
43 
+The input data can be either a Summarized Experiment object of a matrix
44 
+(samples as columns, probes as rows) from the following platforms:
45 
+
46 
+- HM27
47 
+- HM450
48 
+- EPIC array.
49 
+
50 
+In this example we will retrieve two samples from TCGA and classify them expecting
51 
+the same result as the paper.
52 
+
53 
+```{r, eval = TRUE, message = FALSE, results = "hide"}
54 
+query <- GDCquery(project = "TCGA-GBM",
55 
+                   data.category = "DNA methylation",
56 
+                   barcode = c("TCGA-06-0122","TCGA-14-1456"),
57 
+                   platform = "Illumina Human Methylation 27",
58 
+                   legacy = TRUE)
59 
+GDCdownload(query)
60 
+data.hg19 <- GDCprepare(query)
61 
+```
62 
+
63 
+```{r, eval = TRUE}
64 
+assay(data.hg19)[1:5,1:2]
65 
+```
66 
+
67 
+### Function
68 
+```{r, eval = TRUE}
69 
+classification <- gliomaClassifier(data.hg19)
70 
+```
71 
+
72 
+### Results
73 
+
74 
+The classfier will return a list of 3 data frames:
75 
+
76 
+1. Sample final classification
77 
+2. Each model final classification
78 
+3. Each class probability of classification
79 
+
80 
+```{r, eval = TRUE}
81 
+names(classification)
82 
+classification$final.classification
83 
+classification$model.classifications
84 
+classification$model.probabilities
85 
+```
86 
+
87 
+### Comparing results with paper
88 
+```{R}
89 
+TCGAquery_subtype("GBM") %>%
90 
+ dplyr::filter(patient %in% c("TCGA-06-0122","TCGA-14-1456")) %>%
91 
+ dplyr::select("patient","Supervised.DNA.Methylation.Cluster")
92 
+```
93 
+