accio/bioqc

Detect tissue heterogeneity in gene expression data with BioQC

bioqc's Issues

Update BioQC on Bioconductor to the latest version.

The version on Bioconductor is outdated (old version that uses a different API for the wmwTest function.)

Command-line script for BioQC

For integrating BioQC in automated workflows (such as nfcore/rnaseq) it would be really helpful to have a CLI to BioQC.
Essentially, the script should take

• a gene expression matrix
• optionally, a GMT file, or just use BioQC's default if none is specified
• optionally, a species identifier to automatically remap to orthologous gene symbols (maybe even autodetect that for rat and mouse)

and produce

• A heatmap with all signatures that are detected in at least one sample
• A CSV/TSV file with all signature scores

Readme should link to github pages

http://accio.github.io/BioQC/

SignedGenesets gives `NaN`s for `valType %in% c("r", "f")`

Hi @Accio ,

I noticed that wmwTest() on SignedGenesets returns NaNs for valTypes "r" and "f". Not sure if that is intentional.

m <- matrix(c(
  0, 0, 0, 0, 0, 5,
  0, 0, 0, 0, 5, 5,
  0, 0, 0, 5, 5, 5,
  5, 5, 0, 0, 0, 0,
  5, 0, 0, 0, 0, 0
), byrow=TRUE, nrow=5)

rownames(m) <- paste0("g", 1:5)
colnames(m) <- paste0("s", 1:6)

sign <- BioQC::SignedGenesets(list(
  list(name="grows left to right", positive = c("g1", "g2", "g3"), negative = c()),
  list(name="grows left to right with neg", positive = c("g1", "g2", "g3"), negative = c("g4", "g5"))
))

BioQC::wmwTest(m, sign, valType="U1")
#>                              s1   s2 s3 s4 s5 s6
#> grows left to right           0  1.5  3  4  5  6
#> grows left to right with neg -6 -3.0  0  2  4  6
BioQC::wmwTest(m, sign, valType="r")
#>                                s1   s2  s3        s4        s5  s6
#> grows left to right            -1 -0.5   0 0.3333333 0.6666667   1
#> grows left to right with neg -Inf -Inf NaN       Inf       Inf Inf
BioQC::wmwTest(m, sign, valType="f")
#>                                s1   s2  s3        s4        s5  s6
#> grows left to right             0 0.25 0.5 0.6666667 0.8333333   1
#> grows left to right with neg -Inf -Inf NaN       Inf       Inf Inf

^{Created on 2021-08-12 by the reprex package (v2.0.1)}

Sentence describing performance in documentation does not make sense.

wmwTest.rd, Line 10
https://github.com/Accio/BioQC/blob/master/man/wmwTest.Rd#L10

MultiQC module for BioQC

Once we have a CLI (#32) the logical next step would be to make a MultiQC module to embed BioQC results into multiqc reports.

U value inverted?

Hi David,

I noticed that U value computed by wmwTest seems to be inverted. This also results in production of inverted r and f values.

Please find a reprex below.

What do you think is the right solution for this?

1. Update only behavior of r and r scores?
2. Also update the U score computation and potentially break compatibility? E.g., with some warning messages?
3. Add a flag inverted_U=TRUE, which is default?
4. Something else?

library(BioQC)
#> Loading required package: Biobase
#> Loading required package: BiocGenerics
#> Loading required package: parallel
#> 
#> Attaching package: 'BiocGenerics'
#> The following objects are masked from 'package:parallel':
#> 
#>     clusterApply, clusterApplyLB, clusterCall, clusterEvalQ,
#>     clusterExport, clusterMap, parApply, parCapply, parLapply,
#>     parLapplyLB, parRapply, parSapply, parSapplyLB
#> The following objects are masked from 'package:stats':
#> 
#>     IQR, mad, sd, var, xtabs
#> The following objects are masked from 'package:base':
#> 
#>     anyDuplicated, append, as.data.frame, basename, cbind, colnames,
#>     dirname, do.call, duplicated, eval, evalq, Filter, Find, get, grep,
#>     grepl, intersect, is.unsorted, lapply, Map, mapply, match, mget,
#>     order, paste, pmax, pmax.int, pmin, pmin.int, Position, rank,
#>     rbind, Reduce, rownames, sapply, setdiff, sort, table, tapply,
#>     union, unique, unsplit, which, which.max, which.min
#> Welcome to Bioconductor
#> 
#>     Vignettes contain introductory material; view with
#>     'browseVignettes()'. To cite Bioconductor, see
#>     'citation("Biobase")', and for packages 'citation("pkgname")'.
set.seed(10)
x <- rnorm(10, 1)
y <- rnorm(100, 0)

wilcox.test(x, y, alternative='greater')
#> 
#>  Wilcoxon rank sum test with continuity correction
#> 
#> data:  x and y
#> W = 697, p-value = 0.02052
#> alternative hypothesis: true location shift is greater than 0
# p.value is computed correctly
wmwTest(c(x, y), seq_along(x), valType='p.greater')
#> [1] 0.02052079
# U/W value seems to be inverted
wmwTest(c(x, y), seq_along(x), valType='U')
#> [1] 303
# let's invert and check
wmwTest(c(y, x), seq_along(y), valType='U')
#> [1] 697

^{Created on 2020-12-07 by the reprex package (v0.3.0)}

Update default gmt file to the latest version

The gmt file shipped with BioQC ist more than three years old and has been improved since then.

include `valType=c("r", "f")` into documentation

Hi @Accio ,

I noticed that documentation was not updated to include new wmwTest scores (r and f).

fix in #29

Best,
@idavydov

Update author list in Supplementary docs.

BioQC::gini cannot handle negative values

By definition, gini can only be calculated on positive values.
The problem can be fixed by shifting/normalizing values.
Alternatively, an error message should be thrown. Atm. a nonsense value is calculated.

BioQC::gini should address NA properly

When running BioQC on an expression set, take only rows with gene symbol as background.

because of the reasons discussed...

I'm not sure what should be the default behavior, though. But taking all genes as background could result in misleading results and significant 'over-estimation' of signatures.

wmwTest does not accept default value for simplify with datatypes ("matrix" , "indexList")

(issue was accidentically closed by specifying wrong issue id in commit)

wmwTest(eset, gmt, valType="p.greater") does work (see https://github.com/Accio/BioQC-example/blob/master/bioqc-kidney.Rnw#L94)

whereas

wmwTest(mat, tissueInds, valType="p.greater") does not (see https://github.com/Accio/BioQC-example/blob/master/bioqc-simulation.Rnw#L149)

Problem with gmt files importet from broadinstitute

I am trying to use a gmt file from broad with wmwTest instead of the gmt file included in the package, but I get Error in FUN(X[[i]], ...) : index must be either integer vector in R. I tried to ways of importing the gmt file as below, none were successful.

gmt <- Sys.glob("E:/files/Rspace/geneset.gmt")
gmt<- read.gmt(gmt)
and
gmt <- Sys.glob("E:/files/Rspace/geneset.gmt")
gmt<- getGmt(gmt)

gmt files can be found here http://software.broadinstitute.org/gsea/downloads.jsp
Any solution?

wmwTest does not accept default value for simplify with datatypes ("matrix" , "indexList")

wmwTest(eset, gmt, valType="p.greater") does work (see https://github.com/Accio/BioQC-example/blob/master/bioqc-kidney.Rnw#L94)

whereas

wmwTest(mat, tissueInds, valType="p.greater") does not (see https://github.com/Accio/BioQC-example/blob/master/bioqc-simulation.Rnw#L149)

Use of BioQC package to compute Shannon entropy on transcriptomic data

Hi,

I get really interested by your package as you integrated the metrics included in your paper Martinez and Reyes-Valdès PNAS 2008 (shannon index, diversity and specificity of gene expression). Would you have any tutorial to show how to compute adapted Shannon entropy, gene specificity, specialization of tissues as you define and apply them in your paper ?

Thanks,
Lucile

subsetting a non-existant signature should return an error?

Hi @Accio ,

I noticed that the following code works (and returns an invalid GmtList).

suppressPackageStartupMessages(library(BioQC))
gmtFile <- system.file("extdata/exp.tissuemark.affy.roche.symbols.gmt", package="BioQC")
gmt <- readGmt(gmtFile)
gmt[c("Adipose_NGS_RNASEQATLAS_0.6_3", "NONEXISTeNT", "Brain_NGS_RNASEQATLAS_0.6_3")]
#> A gene-set list in GMT format with 3 genesets
#> Gene-sets:
#>   Adipose_NGS_RNASEQATLAS_0.6_3 (Roche,n=28): THRSP,KRT14,APOB,...
#> character(0)
#>   Brain_NGS_RNASEQATLAS_0.6_3 (Roche,n=387): MIR219-2,MIR378B,SLC17A6,...

^{Created on 2022-01-25 by the reprex package (v2.0.1)}

Sharing this in case you would like to add some kind of warning in this case. Please feel free to close in case you think it's an expected behavior.

names(GmtList) will remove names

Reported by I. Davydov

library(BioQC)
gmt_file <- system.file("extdata/exp.tissuemark.affy.roche.symbols.gmt", package="BioQC")
names(readGmt(gmt_file))
names(GmtList(readGmt(gmt_file)))

Check issues

• checking for missing documentation entries ... WARNING
  Undocumented code objects:
  ‘IndexList’ ‘SignedIndexList’ ‘matchGenes’ ‘offset<-’ ‘valTypes’
  Undocumented S4 classes:
  ‘SignedIndexList’
  Undocumented S4 methods:
  generic 'SignedIndexList' and siglist 'list'
  generic 'matchGenes' and siglist 'GmtList,character'
  generic 'matchGenes' and siglist 'GmtList,eSet'
  generic 'matchGenes' and siglist 'GmtList,matrix'
  generic 'matchGenes' and siglist 'SignedGenesets,character'
  generic 'matchGenes' and siglist 'SignedGenesets,eSet'
  generic 'matchGenes' and siglist 'SignedGenesets,matrix'
  generic 'matchGenes' and siglist 'character,character'
  generic 'matchGenes' and siglist 'character,eSet'
  generic 'matchGenes' and siglist 'character,matrix'
  generic 'offset<-' and siglist 'IndexList,numeric'
  generic 'offset<-' and siglist 'SignedIndexList,numeric'
  generic 'wmwTest' and siglist 'ANY,list'
  generic 'wmwTest' and siglist 'ANY,logical'
  generic 'wmwTest' and siglist 'ANY,numeric'
  generic 'wmwTest' and siglist 'eSet,GmtList'
  generic 'wmwTest' and siglist 'eSet,SignedIndexList'
  generic 'wmwTest' and siglist 'eSet,list'
  generic 'wmwTest' and siglist 'eSet,logical'
  generic 'wmwTest' and siglist 'eSet,numeric'
  generic 'wmwTest' and siglist 'matrix,GmtList'
  generic 'wmwTest' and siglist 'matrix,IndexList'
  generic 'wmwTest' and siglist 'matrix,SignedIndexList'
  generic 'wmwTest' and siglist 'numeric,IndexList'
  generic 'wmwTest' and siglist 'numeric,SignedIndexList'

• checking for code/documentation mismatches ... WARNING
  Codoc mismatches from documentation object 'wmwTest':
  wmwTest
  Code: function(object, indexList, ...)
  Docs: function(x, ind.list, valType = c("p.greater", "p.less",
  "p.two.sided", "U", "abs.log10p.greater",
  "log10p.less", "abs.log10p.two.sided", "Q"), simplify
  = TRUE)
  Argument names in code not in docs:
  object indexList ...
  Argument names in docs not in code:
  x ind.list valType simplify
  Mismatches in argument names:
  Position: 1 Code: object Docs: x
  Position: 2 Code: indexList Docs: ind.list
  Position: 3 Code: ... Docs: valType

• checking Rd \usage sections ... WARNING
  Undocumented arguments in documentation object 'IndexList,list-method'
  ‘keepNA’ ‘keepDup’

Undocumented arguments in documentation object 'IndexList,logical-method'
‘keepNA’ ‘keepDup’

Undocumented arguments in documentation object 'IndexList,numeric-method'
‘keepNA’ ‘keepDup’

Undocumented arguments in documentation object 'as.gmtlist'
‘list’ ‘description’
Documented arguments not in \usage in documentation object 'as.gmtlist':
‘list:’ ‘description:’

Undocumented arguments in documentation object 'matchGenes.signedDefault'
‘signedGenesets’
Documented arguments not in \usage in documentation object 'matchGenes.signedDefault':
‘gmtList’

Undocumented arguments in documentation object 'parseIndex'
‘x’
Documented arguments not in \usage in documentation object 'parseIndex':
‘list’ ‘offset’

Undocumented arguments in documentation object 'readSignedGmt'
‘posPattern’ ‘negPattern’ ‘nomatch’
Documented arguments not in \usage in documentation object 'readSignedGmt':
‘posPattern:’ ‘negPattern:’ ‘nomatch:’

Functions with \usage entries need to have the appropriate \alias
entries, and all their arguments documented.
The \usage entries must correspond to syntactically valid R code.
See chapter ‘Writing R documentation files’ in the ‘Writing R
Extensions’ manual.

00check-20161108.txt

Error when running

Hello,

I'm getting this error when running from command line

wmwTest(x = t(data), indexList = (trait==names(table(trait))[1]), valType="p.two.sided", simplify = TRUE).

I don't get the error when I run from Rstudio. The sessioninfo seems to be the same. Have you seen this before ?

Error in .setupMethodsTables(fdef, initialize = TRUE) :
  trying to get slot "group" from an object of a basic class ("NULL") with no slots
Calls: filterfeaturesK ... validityMethod -> as -> .getMethodsTable -> .setupMethodsTables
Execution halted

> sessionInfo()
R version 3.4.3 (2017-11-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS: /share/apps/R-latest/lib64/R/lib/libRblas.so
LAPACK: /share/apps/R-latest/lib64/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
[1] BioQC_1.6.0         Biobase_2.38.0      BiocGenerics_0.24.0
[4] Rcpp_0.12.15

loaded via a namespace (and not attached):
[1] compiler_3.4.3 tools_3.4.3