Hello! I got such an error report when I include a bed file with no peak into my DBA sample sheet.
Error in mergeScores(merged, def, peakset, TRUE) : Not compatible with requested type: [type=character; target=double]
Primarily, I do not want to delete this 0-byte bed file from the sample sheet, because its parallel replicates still has some peaks. But if I have to delete it (replicate no is 2), should I also change other replicate numbers, for example, 4 to 3 and 3 to 2? What is the best solution? Thank you very much!
Thank you Dr. Stark. I tried to construct the DBA object with the data sheet containing one "empty" value at Peaks. Unfortunately, the error was still there. Then I replaced that peaks value into a manually-created bed file containing only one peak and zero strength. This time the data sheet can be successfully loaded. Is this manually-created bed file good for further analysis. And of course I still want to know why the "empty" setting got problem.
This question becomes important to me now again because I plan to fill some rows with the files of "IgG vs IgG". My previous rows linked to the files of "Scramble_antibody vs Scramble_IgG" and "shRNA_antibody vs shRNA_IgG". Here "vs" means the latter one's strength was taken as background in peak caller. I tried plotVenn() to see the overlap of "Scramble_antibody vs Scramble_IgG" and "shRNA_antibody vs shRNA_IgG", but the setting with only one contrast does not meet the criteria. So I considered adding IgG samples as new rows in DBA to generate 2 contrasts. On the other hand, in the peak caller SEACR I used, IgG is always recommended to keep the threshold setting. As a solution, though it sounds strange, I tried to generate a "IgG vs IgG" bed file in SEACR. And the online SEACR just froze on my request. So as a final solution, I came back to the idea of "muting" the peaks value in new rows by setting "empty" or manually filling with a zero-strength and one-peak file. I hope my description makes sense and wanna your suggestion. Maybe there is a simpler way.
Thank you very much!
Can you send along the output of
sessionInfo()
so I can see what versions you are working with?Dr. Stark, here please see what returned from sessionInfo(). Thank you!
R version 4.0.3 (2020-10-10) Platform: x86_64-apple-darwin17.0 (64-bit) Running under: macOS High Sierra 10.13.6
Matrix products: default BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib
locale: [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
attached base packages: [1] parallel stats4 stats graphics grDevices utils datasets [8] methods base
other attached packages: [1] DiffBind_3.0.7 SummarizedExperiment_1.20.0 [3] Biobase_2.50.0 MatrixGenerics_1.2.0
[5] matrixStats_0.57.0 GenomicRanges_1.42.0
[7] GenomeInfoDb_1.26.1 IRanges_2.24.0
[9] S4Vectors_0.28.0 BiocGenerics_0.36.0
loaded via a namespace (and not attached): [1] backports_1.2.0 GOstats_2.56.0 BiocFileCache_1.14.0
[4] plyr_1.8.6 GSEABase_1.52.0 splines_4.0.3
[7] BiocParallel_1.24.1 ggplot2_3.3.2 amap_0.8-18
[10] digest_0.6.27 invgamma_1.1 GO.db_3.12.1
[13] SQUAREM_2020.5 magrittr_2.0.1 checkmate_2.0.0
[16] memoise_1.1.0 BSgenome_1.58.0 base64url_1.4
[19] limma_3.46.0 Biostrings_2.58.0 annotate_1.68.0
[22] systemPipeR_1.24.2 askpass_1.1 bdsmatrix_1.3-4
[25] prettyunits_1.1.1 jpeg_0.1-8.1 colorspace_2.0-0
[28] blob_1.2.1 rappdirs_0.3.1 apeglm_1.12.0
[31] ggrepel_0.8.2 dplyr_1.0.2 crayon_1.3.4
[34] RCurl_1.98-1.2 jsonlite_1.7.1 graph_1.68.0
[37] genefilter_1.72.0 brew_1.0-6 survival_3.2-7
[40] VariantAnnotation_1.36.0 glue_1.4.2 gtable_0.3.0
[43] zlibbioc_1.36.0 XVector_0.30.0 DelayedArray_0.16.0
[46] V8_3.4.0 Rgraphviz_2.34.0 scales_1.1.1
[49] pheatmap_1.0.12 mvtnorm_1.1-1 DBI_1.1.0
[52] edgeR_3.32.0 Rcpp_1.0.5 xtable_1.8-4
[55] progress_1.2.2 emdbook_1.3.12 bit_4.0.4
[58] rsvg_2.1 AnnotationForge_1.32.0 truncnorm_1.0-8
[61] httr_1.4.2 gplots_3.1.1 RColorBrewer_1.1-2
[64] ellipsis_0.3.1 pkgconfig_2.0.3 XML_3.99-0.5
[67] dbplyr_2.0.0 locfit_1.5-9.4 tidyselect_1.1.0
[70] rlang_0.4.9 AnnotationDbi_1.52.0 munsell_0.5.0
[73] tools_4.0.3 generics_0.1.0 RSQLite_2.2.1
[76] stringr_1.4.0 yaml_2.2.1 bit64_4.0.5
[79] caTools_1.18.0 purrr_0.3.4 RBGL_1.66.0
[82] xml2_1.3.2 biomaRt_2.46.0 compiler_4.0.3
[85] rstudioapi_0.13 curl_4.3 png_0.1-7
[88] geneplotter_1.68.0 tibble_3.0.4 stringi_1.5.3
[91] GenomicFeatures_1.42.1 lattice_0.20-41 Matrix_1.2-18
[94] vctrs_0.3.5 pillar_1.4.7 lifecycle_0.2.0
[97] data.table_1.13.2 bitops_1.0-6 irlba_2.3.3
[100] rtracklayer_1.50.0 R6_2.5.0 latticeExtra_0.6-29
[103] hwriter_1.3.2 ShortRead_1.48.0 KernSmooth_2.23-18
[106] MASS_7.3-53 gtools_3.8.2 assertthat_0.2.1
[109] DESeq2_1.30.0 openssl_1.4.3 Category_2.56.0
[112] rjson_0.2.20 withr_2.3.0 GenomicAlignments_1.26.0 [115] batchtools_0.9.14 Rsamtools_2.6.0 GenomeInfoDbData_1.2.4
[118] hms_0.5.3 grid_4.0.3 DOT_0.1
[121] coda_0.19-4 GreyListChIP_1.22.0 ashr_2.2-47
[124] mixsqp_0.3-43 bbmle_1.0.23.1 numDeriv_2016.8-1.1
In general you do not pass contrasts to
dba.plotVenn()
, you pass a mask designating 2, 3 or 4 samples. If you have two conditions with replicates, you probably need to make a consensus for each conditions first, then show their overlap, eg:If you include output of the DBA object (what print out if you just type the name of the object), I can be more specific.
Thank you so much Dr. Stark. I am pasting my plans of Venn diagrams as below.
Plan1:
Venn1, is the venn diagram of KWScr, KWsh5, NWScr_sb_sh5 and MWScr_sb_sh5.
Venn2, is the venn diagram of KKScr, KKsh5, NKScr_sb_sh5 and MKScr_sb_sh5.
Venn3, is the Venn diagram of commen peaks from Venn1 and Venn2.
Plan2 (just add some concerns on IgG into Plan1):
Venn1, is the venn diagram of 3 constrasts. Contrast1=KWScr vs RWScr , Contrast2=KWsh5 vs RWsh5, Contrast3=NWScr_sb_sh5 vs MWScr_sb_sh5.
Venn2, is the venn diagram of 3 constrasts. Contrast4=KKScr vs RKScr , Contrast5=KKsh5 vs RKsh5, Contrast6=NKScr_sb_sh5 vs MKScr_sb_sh5.
Venn3, is the Venn diagram of commen peaks from Venn1 and Venn2.
Venn4, is the Venn diagram of Contrasts1-6.
And I found
dba.plotVenn(anaKRWnmlwcst2,contrast=1:2,method=DBA_ALL_METHODS_BLOCK)
works butdba.plotVenn(anaKRWnmlwcst2,contrast=1:2:3,method=DBA_ALL_METHODS_BLOCK)
ordba.plotVenn(anaKRWnmlwcst2,contrast=1:2,1:3,2:3,method=DBA_ALL_METHODS_BLOCK)
does not work.I have hitherto found
contrast=c(1,2,3)
works. Cheers! But meanwhile I found if more rows were added to DBA to generate new plotvenn, those previous low peak numbers would be recalculated and even to zero. Could I generate Venn diagram from more than one DBA? BTW, I also trieddba.peakset()
, but it showed "DBA object is already formed from a consensus peakset!" andpdba.plotVenn(anaKRNMWnmlwcst3, anaKRNMWnmlwcst3$masks$Consensus)
gave "error Too many peaksets in mask." I will retry without contrasts steps.