|
...
|
...
|
@@ -76,16 +76,30 @@
|
|
76
|
76
|
</div>
|
|
77
|
77
|
</nav><div class="container template-home">
|
|
78
|
78
|
<div class="row">
|
|
79
|
|
- <main id="main" class="col-md-9">
|
|
80
|
|
-The software formalises a framework for classification in R.
|
|
81
|
|
- There are four stages; Data transformation, feature selection, classifier training,
|
|
82
|
|
- and prediction. The requirements of variable types and names are
|
|
83
|
|
- fixed, but specialised variables for functions can also be provided.
|
|
84
|
|
- The classification framework is wrapped in a driver loop, that
|
|
85
|
|
- reproducibly carries out a number of cross-validation schemes.
|
|
86
|
|
- Functions for differential expression, differential variability,
|
|
87
|
|
- and differential distribution are included. Additional functions
|
|
88
|
|
- may be developed by the user, by creating an interface to the framework.
|
|
|
79
|
+ <main id="main" class="col-md-9"><div class="section level1">
|
|
|
80
|
+<div class="page-header"><h1 id="classifyr-performance-evaluation-for-multi-view-data-sets-and-seamless-integration-with-multiassayexperiment-and-bioconductor">ClassifyR: Performance evaluation for multi-view data sets and seamless integration with MultiAssayExperiment and Bioconductor<a class="anchor" aria-label="anchor" href="#classifyr-performance-evaluation-for-multi-view-data-sets-and-seamless-integration-with-multiassayexperiment-and-bioconductor"></a>
|
|
|
81
|
+</h1></div>
|
|
|
82
|
+<p><img src="reference/figures/ClassifyRsticker.png" align="right"></p>
|
|
|
83
|
+<p>ClassifyR’s performance evaluation focuses on model stability and interpretability. Based on repeated cross-validation, it is possible to evaluate feature selection stability and also per-sample prediction accuracy. Also, multiple omics data assays on the same samples are becoming more popular and ClassifyR supports a range of multi-view methods to evaluate which data view is the most predictive and combine data views to evaluate if multiple views provide superior predictive performance to a single data view.</p>
|
|
|
84
|
+<div class="section level2">
|
|
|
85
|
+<h2 id="installation">Installation<a class="anchor" aria-label="anchor" href="#installation"></a>
|
|
|
86
|
+</h2>
|
|
|
87
|
+<p>The recommended method of installing ClassifyR is by using Bioconductor’s BiocManager installer:</p>
|
|
|
88
|
+<pre><code><span><span class="kw"><a href="https://rdrr.io/r/base/library.html" class="external-link">library</a></span><span class="op">(</span><span class="va">BiocManager</span><span class="op">)</span></span>
|
|
|
89
|
+<span><span class="fu"><a href="https://rdrr.io/pkg/BiocManager/man/install.html" class="external-link">install</a></span><span class="op">(</span><span class="st">"ClassifyR"</span>, dependencies <span class="op">=</span> <span class="cn">TRUE</span><span class="op">)</span></span></code></pre>
|
|
|
90
|
+<p>The above code will install all packages that provide feature selection or model-building functionality. If only one or two methods are desired then the dependencies option could be omitted and those packages providing functionality installed manually.</p>
|
|
|
91
|
+</div>
|
|
|
92
|
+<div class="section level2">
|
|
|
93
|
+<h2 id="website">Website<a class="anchor" aria-label="anchor" href="#website"></a>
|
|
|
94
|
+</h2>
|
|
|
95
|
+<p>Please visit <a href="https://sydneybiox.github.io/ClassifyR/">the ClassifyR website</a> to view the main vignette as well as articles that provide more in-depth explanations for various aspects of the package. Details of performance evaluation, multi-view methods and contributing a wrapper for a new algorithm to the package are provided.</p>
|
|
|
96
|
+</div>
|
|
|
97
|
+<div class="section level2">
|
|
|
98
|
+<h2 id="reference">Reference<a class="anchor" aria-label="anchor" href="#reference"></a>
|
|
|
99
|
+</h2>
|
|
|
100
|
+<p>Strbenac D., Mann, G.J., Ormerod, J.T., and Yang, J. Y. H. (2015) ClassifyR: An R package for performance assessment of classification with applications to transcriptomics, <em>Bioinformatics</em>.</p>
|
|
|
101
|
+</div>
|
|
|
102
|
+</div>
|
|
89
|
103
|
</main><aside class="col-md-3"><div class="links">
|
|
90
|
104
|
<h2 data-toc-skip>Links</h2>
|
|
91
|
105
|
<ul class="list-unstyled">
|
