Dual-flashlight plot: Difference between revisions

Revision as of 00:30, 28 November 2018

Latest revision as of 03:53, 3 December 2023

(3 intermediate revisions by 3 users not shown)

Line 1:

[[Image:wiki.dualflashlight.png|thumb|350px|Dual-flashlight plot showing a high-throughput screening dataset.]]

In statistics, a '''dual-flashlight plot''' is a type of scatter-plot in which the standardized mean of a contrast variable ([[SMCV]]) is plotted against the mean of a [[contrast variable]] representing a comparison of interest

.<ref name=ZhangPharmacogenomics2010>{{cite journal |author=Zhang XHD

|title= Assessing the size of gene or RNAi effects in multifactor high-throughput experiments

|journal=Pharmacogenomics |volume=11 |issue= 2|pages=199–213

|year=2010 |month= |pmid= 20136359|doi=10.2217/PGS.09.136 |url=}}</ref> The commonly used dual-flashlight plot is for the difference between two groups in high-throughput experiments such as [[microarray]]s and [[high-throughput screening]] studies, in which we plot the [[SSMD]] versus average log fold-change on the ''y''- and ''x''-axes, respectively, for all genes or compounds (such as [[siRNA]]s or [[small molecule]]s) investigated in an experiment.<ref name="ZhangPharmacogenomics2010"/>

As a whole, the points in a dual-flashlight plot look like the beams of a flashlight with two heads, hence the name dual-flashlight plot.<ref name="ZhangPharmacogenomics2010"/>

With the dual-flashlight plot, we can see how the genes or compounds are distributed into each category in effect sizes, as shown in the figure. Meanwhile, we can also see the average fold-change for each gene or compound. The dual-flashlight plot is similar to the [[volcano plot (statistics)|volcano plot]]. In a [[volcano plot (statistics)|volcano plot]], the [[p-value]] (or q-value{{clarify|date=June 2012}}), instead of SMCV or SSMD, is plotted against average fold-change <ref name=Jin2001>{{cite journal |vauthors=Jin W, Riley RM, Wolfinger RD, White KP, Passador-Gurgel G, Gibson G |title= The contributions of sex, genotype and age to transcriptional variance in Drosophila melanogaster

|journal= Nature Genetics |volume=29 |issue= 4|pages=389–95

|year=2001 |month= |pmid= 11726925|doi= 10.1038/ng766 |urls2cid= 16841881
}}</ref>

.<ref name=Cui2003>{{cite journal |vauthors=Cui X, Churchill GA |title=Statistical tests for differential expression in cDNA microarray experiments

|journal= Genome Biology |volume=4 |issue= 4|pages=210

|year=2003 |month= |pmid=12702200 |doi= 10.1186/gb-2003-4-4-210|urlpmc= 154570
|pmcdoi-access=154570free
}}</ref> The advantage of using SMCV over p-value (or q-value) is that, if there exist any non-zero true effects for a gene or compound, the estimated SMCV goes to its population value whereas the p-value (or q-value) for testing no mean difference (or zero contrast mean) goes to zero when the sample size increases

.<ref name=ZhangSBR2010>{{cite journal |author=Zhang XHD

|title= Strictly standardized mean difference, standardized mean difference and classical t-test for the comparison of two groups

|journal= Statistics in Biopharmaceutical Research |volume=2 |issue= 2|pages=292–99

|year=2010 |doi=10.1198/sbr.2009.0074 |s2cid= 119825625

|year=2010 |month= |pmid= |doi=10.1198/sbr.2009.0074 |url=}}</ref> Hence, the value of SMCV is comparable whereas the value of p-value or q-value is not comparable in experiments with different sample size, especially when many investigated genes or compounds do not have exactly zero effects. The dual-flashlight plot bears the same advantage that the SMCV has, as compared to the [[volcano plot (statistics)|volcano plot]].

==See also==