0
|
1 <tool id="secimtools_multiple_testing_adjustment" name="Multiple Testing Adjustment (MTA)" version="@WRAPPER_VERSION@">
|
|
2 <description>of p-values.</description>
|
|
3 <macros>
|
|
4 <import>macros.xml</import>
|
|
5 </macros>
|
|
6 <expand macro="requirements" />
|
|
7 <command><![CDATA[
|
|
8 multiple_testing_adjustment.py
|
|
9 --input $input
|
|
10 --uniqID $uniqID
|
|
11 --pval "$pval"
|
|
12 --alpha $alpha
|
|
13 --outadjusted $outadjusted
|
|
14 --flags $flags
|
|
15 ]]></command>
|
|
16 <inputs>
|
|
17 <param name="input" type="data" format="tabular" label="Wide Dataset" help="Input your tab-separated wide format dataset. If not tab separated see TIP below."/>
|
|
18 <param name="uniqID" type="text" size="30" value="" label="Unique Feature ID" help="Name of the column in your Wide Dataset that has unique identifiers.."/>
|
|
19 <param name="pval" type="text" size="30" value="" label="p-value column" help="Name of the column in your wide dataset that contains the p-values."/>
|
|
20 <param name="alpha" type="float" size="6" value="0.05" label="α" help="Value of α to be used for multiple correction. Default α = 0.05."/>
|
|
21 </inputs>
|
|
22 <outputs>
|
|
23 <data format="tabular" name="outadjusted" label="${tool.name} on ${on_string}: Adjusted pval."/>
|
|
24 <data format="tabular" name="flags" label="${tool.name} on ${on_string}: Flags."/>
|
|
25 </outputs>
|
|
26 <tests>
|
|
27 <test>
|
|
28 <param name="input" value="ST000006_anova_fixed_with_group_summary.tsv"/>
|
|
29 <param name="uniqID" value="Retention_Index" />
|
|
30 <param name="pval" value="prob_greater_than_t_for_diff_Chardonnay, Carneros, CA 2003 (CH01)-Chardonnay, Carneros, CA 2003 (CH02)" />
|
|
31 <param name="alpha" value="0.05" />
|
|
32 <output name="outadjusted" file="ST000006_multiple_testing_adjustment_outadjusted.tsv" />
|
|
33 <output name="flags" file="ST000006_multiple_testing_adjustment_flags.tsv" />
|
|
34 </test>
|
|
35 </tests>
|
|
36 <help><![CDATA[
|
|
37
|
|
38 @TIP_AND_WARNING@
|
|
39
|
|
40 **Tool Description**
|
|
41
|
|
42 The tool is designed to adjust p-values for multiple comparisons using three different methods:
|
|
43
|
|
44 (1) The Bonferroni method and two false discovery rate (FDR) methods, (2) the Benjamini-Hochberg method (BH) and (3) the Benjamini-Yekutieli method (BY).
|
|
45 The p-value correction can be carried out on p-values generated from the following tools: Analysis of Variance (ANOVA) Fixed Effects Model, Kruskal-Wallis Non-Parametric Test, T-test (Single Group) and T-test (Paired and/or Unpaired) in addition to p-values generated outside of these tools.
|
|
46 The user can specify the total type I error α value.
|
|
47
|
|
48 More details about the PH and BY methods are available in the papers:
|
|
49
|
|
50 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the royal statistical society. Series B (Methodological), 289-300.
|
|
51
|
|
52 Benjamini, Y., and Yekutieli, D. (2001). The control of the false discovery rate in multiple testing under dependency. Annals of statistics, 1165-1188.
|
|
53
|
|
54 -------------------------------------------------------------------------------------------
|
|
55
|
|
56 **Input**
|
|
57
|
|
58 - Two input datasets are required.
|
|
59
|
|
60 @WIDE@
|
|
61
|
|
62 @UNIQID@
|
|
63
|
|
64 **Name for p-value column**
|
|
65
|
|
66 - Name of the column in your Wide Dataset that contains the p-values.
|
|
67
|
|
68 **α**
|
|
69
|
|
70 - Value of α to be used for multiple correction. Default α = 0.05.
|
|
71
|
|
72 -------------------------------------------------------------------------------------------
|
|
73
|
|
74 **Output**
|
|
75
|
|
76 The tool produces two TSV files:
|
|
77
|
|
78 (1) One TSV that contains the following five columns:
|
|
79 a column with unique feature IDs,
|
|
80 a column of the original p-values and
|
|
81 the last three columns contain the p-values adjusted using the 3 methods described above which are reflected in the column name.
|
|
82 (2) The second TSV file contains flags where all significant values are flagged as 1 and non-significant values are flagged as 0.
|
|
83
|
|
84 ]]></help>
|
|
85 <expand macro="citations"/>
|
|
86 </tool>
|