DE analysis between conditions (tissue)

Code

• Show All Code
• Hide All Code
• Download Rmd

Eed-cKO vs WT comparison AP-axis

DE for Eed-cKO vs WT on FB, MB, HB, SC

In this notebook we test for differential expression between and WT tissues in the developing mouse CNS.

Each condition (WT, , for FB, MB, HB, and SC) contains six samples (two replicates at e13.5, e15.5, and e18).

For this analysis, we are not interested in the effects of time and thus add in the time factor to our DE (linear) model as a batch correction factor.

Script setup

Run DE on each tissue

Here we run the DE analysis on each tissue for the comparison.

runDeseq2BetweenCond(paste('merged_df_fb_FEATURE_COUNTS_', date, '.csv', sep=''), paste('DEseq2_CNS_fb_', date, '.csv', sep=''))

[1] "========================== RUNNING merged_df_fb_FEATURE_COUNTS_20210124.csv ============================"
[1] "Dataset dimensions:  15304 12"

estimating size factors
estimating dispersions
gene-wise dispersion estimates

Normalise the data

filename <- paste(project_dir, 'merged_df_FEATURE_COUNTS_', date, '.csv', sep='')
# https://bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html
counts <- read.csv(filename, header = TRUE, sep = ",")
rownames(counts) <- counts$entrezgene_id
genes <- counts$entrezgene_id

# Let's make sure our count data is in matrix format and is only the numeric columns i.e. everything but the genes
counts <- counts[,2:ncol(counts)]
counts[is.na(counts)] = 0
# Here we get the names of the columns, for my stuff I always have all the info in one string as I find it makes it easier
# this means each of the "groups" are separated by a "_" this may be different for you
sampleNames <- colnames(counts) # Sample names

# For DEseq2 we need to turn this into a dataframe 
sampleDF = data.frame(sampleNames = sampleNames)

# Make sure we don't include the ID in our columns
countMatrix <- as.matrix(counts)

# We now set the row names to be the gene IDs
rownames(countMatrix) <- genes

ddsMat <- DESeqDataSetFromMatrix(countData = countMatrix,
                                 colData = sampleDF,
                                 design = ~ 1)

dds <- estimateSizeFactors(ddsMat)
outputFilename <- paste(project_dir, 'merged_df_FEATURE_COUNTS_DEseq2Norm_', date, '.csv', sep='')
write.csv(counts(dds,normalized=TRUE), file = outputFilename)

rld <- rlog(countMatrix + 1, blind = FALSE)

rlog() may take a long time with 50 or more samples,
vst() is a much faster transformation
converting counts to integer mode

Print session info

sessionInfo()

R version 4.0.3 (2020-10-10)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Catalina 10.15.5

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_AU.UTF-8/en_AU.UTF-8/en_AU.UTF-8/C/en_AU.UTF-8/en_AU.UTF-8

attached base packages:
[1] parallel  stats4    stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] edgeR_3.30.3                limma_3.44.3                DESeq2_1.28.1               SummarizedExperiment_1.18.2 DelayedArray_0.14.1        
 [6] matrixStats_0.58.0          Biobase_2.48.0              GenomicRanges_1.40.0        GenomeInfoDb_1.24.2         IRanges_2.22.2             
[11] S4Vectors_0.26.1            BiocGenerics_0.34.0        

loaded via a namespace (and not attached):
 [1] bitops_1.0-7           bit64_4.0.5            RColorBrewer_1.1-2     rstan_2.21.2           tools_4.0.3            bslib_0.2.5            utf8_1.2.1            
 [8] R6_2.5.0               DBI_1.1.1              colorspace_2.0-1       withr_2.4.2            tidyselect_1.1.1       gridExtra_2.3          prettyunits_1.1.1     
[15] processx_3.5.2         sircle_0.0.0.9000      bit_4.0.4              curl_4.3.1             compiler_4.0.3         cli_2.5.0              sass_0.4.0            
[22] scales_1.1.1           genefilter_1.70.0      callr_3.7.0            stringr_1.4.0          digest_0.6.27          StanHeaders_2.21.0-7   rmarkdown_2.8         
[29] XVector_0.28.0         pkgconfig_2.0.3        htmltools_0.5.1.1      fastmap_1.1.0          rlang_0.4.11           RSQLite_2.2.7          jquerylib_0.1.4       
[36] generics_0.1.0         jsonlite_1.7.2         BiocParallel_1.22.0    dplyr_1.0.6            inline_0.3.18          RCurl_1.98-1.3         magrittr_2.0.1        
[43] GenomeInfoDbData_1.2.3 loo_2.4.1              Matrix_1.3-3           Rcpp_1.0.6             munsell_0.5.0          fansi_0.4.2            lifecycle_1.0.0       
[50] stringi_1.5.3          yaml_2.2.1             zlibbioc_1.34.0        pkgbuild_1.2.0         grid_4.0.3             blob_1.2.1             crayon_1.4.1          
[57] lattice_0.20-44        splines_4.0.3          annotate_1.66.0        locfit_1.5-9.4         knitr_1.33             ps_1.6.0               pillar_1.6.1          
[64] geneplotter_1.66.0     codetools_0.2-18       XML_3.99-0.6           glue_1.4.2             evaluate_0.14          V8_3.4.2               RcppParallel_5.1.4    
[71] vctrs_0.3.8            gtable_0.3.0           purrr_0.3.4            assertthat_0.2.1       cachem_1.0.5           ggplot2_3.3.3          xfun_0.23             
[78] xtable_1.8-4           rsconnect_0.8.17       survival_3.2-11        tibble_3.1.0           AnnotationDbi_1.50.3   memoise_2.0.0          ellipsis_0.3.2        

References:

hbctraining/DGE_workshop. Teaching materials at the Harvard Chan Bioinformatics Core, 2021. Accessed: Jun. 15, 2021. [Online]. Available: https://github.com/hbctraining/DGE_workshop

Love MI, Huber W, Anders S (2014). “Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2.” Genome Biology, 15, 550. doi: 10.1186/s13059-014-0550-8.

M. D. Robinson, D. J. McCarthy, and G. K. Smyth, ‘edgeR: a Bioconductor package for differential expression analysis of digital gene expression data’, Bioinformatics, vol. 26, no. 1, pp. 139–140, Jan. 2010, doi: 10.1093/bioinformatics/btp616.

LS0tCnRpdGxlOiAiRWVkLWNLTyB2cyBXVCBjb21wYXJpc29uIEFQLWF4aXMiCm91dHB1dDogaHRtbF9ub3RlYm9vawotLS0KCgpgYGB7ciBzZXR1cCwgaW5jbHVkZT1GQUxTRX0Ka25pdHI6Om9wdHNfY2h1bmskc2V0KGVjaG8gPSBUUlVFKQpgYGAKCiMjIERFIGZvciBFZWQtY0tPIHZzIFdUIG9uIEZCLCBNQiwgSEIsIFNDCgpJbiB0aGlzIG5vdGVib29rIHdlIHRlc3QgZm9yIGRpZmZlcmVudGlhbCBleHByZXNzaW9uIGJldHdlZW4gXGVtcGh7RWVkLWNLT30gYW5kIFdUIHRpc3N1ZXMgaW4gdGhlIGRldmVsb3BpbmcgbW91c2UgQ05TLlwKCkVhY2ggY29uZGl0aW9uIChXVCwgXGVtcGh7RWVkLWNLT30sIGZvciBGQiwgTUIsIEhCLCBhbmQgU0MpIGNvbnRhaW5zIHNpeCBzYW1wbGVzICh0d28gcmVwbGljYXRlcyBhdCBlMTMuNSwgZTE1LjUsIGFuZCBlMTgpLlwKCkZvciB0aGlzIGFuYWx5c2lzLCB3ZSBhcmUgbm90IGludGVyZXN0ZWQgaW4gdGhlIGVmZmVjdHMgb2YgdGltZSBhbmQgdGh1cyBhZGQgaW4gdGhlIHRpbWUgZmFjdG9yIHRvIG91ciBERSAobGluZWFyKSBtb2RlbCBhcyBhIGJhdGNoIGNvcnJlY3Rpb24gZmFjdG9yLlwKCgojIyMgU2NyaXB0IHNldHVwCgpgYGB7cn0KbGlicmFyeShERVNlcTIpCgojIEdsb2JhbCB2YXJpYWJsZXMgdG8gbWFrZSBuYW1lIGZvcm1hdHMgY29uc2lzdGVudApwcm9qZWN0X2RpciA8LSAnLi4vZGF0YS9yZXN1bHRzL2Rlc2VxMi8nCmRhdGUgPC0gJzIwMjEwMTI0JwoKcnVuRGVzZXEyQmV0d2VlbkNvbmQgPC0gZnVuY3Rpb24gKGZpbGVuYW1lLCBvdXRwdXRfZmlsZW5hbWUsIHJ1bl90eXBlKSB7CgogICAgZmlsZV9wYXRoIDwtIHBhc3RlKHByb2plY3RfZGlyLCBmaWxlbmFtZSwgc2VwPScnKQogIAogICAgcHJpbnQocGFzdGUoIj09PT09PT09PT09PT09PT09PT09PT09PT09IFJVTk5JTkcgIiwgZmlsZW5hbWUsICIgPT09PT09PT09PT09PT09PT09PT09PT09PT09PSIsIHNlcD0iIikpCiAgICAgICMgaHR0cHM6Ly9iaW9jb25kdWN0b3Iub3JnL3BhY2thZ2VzL3JlbGVhc2Uvd29ya2Zsb3dzL3ZpZ25ldHRlcy9ybmFzZXFHZW5lL2luc3QvZG9jL3JuYXNlcUdlbmUuaHRtbAogICAgY291bnRzIDwtIHJlYWQuY3N2KGZpbGVfcGF0aCwgaGVhZGVyID0gVFJVRSwgc2VwID0gIiwiKQogICAgcm93bmFtZXMoY291bnRzKSA8LSBjb3VudHMkdV9pZAoKICAgICMgTGV0J3MgbWFrZSBzdXJlIG91ciBjb3VudCBkYXRhIGlzIGluIG1hdHJpeCBmb3JtYXQgYW5kIGlzIG9ubHkgdGhlIG51bWVyaWMgY29sdW1ucyBpLmUuIGV2ZXJ5dGhpbmcgYnV0IHRoZSBnZW5lcwogICAgY291bnRzIDwtIGNvdW50c1ssMjpuY29sKGNvdW50cyldCgogICAgc2FtcGxlX25hbWVzIDwtIGNvbG5hbWVzKGNvdW50cykgIyBTYW1wbGUgbmFtZXMKICAgIAogICAgIyBNYWtlIHN1cmUgd2UgZG9uJ3QgaW5jbHVkZSB0aGUgSUQgaW4gb3VyIGNvbHVtbnMKICAgIGNvdW50X21hdHJpeCA8LSBhcy5tYXRyaXgoY291bnRzKQogICAgCiAgICAjIFdlIG5vdyBzZXQgdGhlIHJvdyBuYW1lcyB0byBiZSB0aGUgZ2VuZSBJRHMKICAgIHJvd25hbWVzKGNvdW50X21hdHJpeCkgPC0gcm93bmFtZXMoY291bnRzKSAKICAgIAogICAgIyBTZXBhcmF0ZSBvdXQgZWFjaCBvZiB0aGUgc2FtcGxlIG5hbWVzIHRvIGJlIHRoZSBkaWZmZXJlbnQgZXhwZXJpbWVudCBjb25kaXRpb25zCiAgICBjb25kaXRpb24gPC0gZmFjdG9yKHNhcHBseShzYW1wbGVfbmFtZXMsIGZ1bmN0aW9uKHgpe3N1YnN0cih4LCBzdGFydCA9IDEsIHN0b3AgPSAyKX0pKQogICAgdGltZSA8LSBmYWN0b3Ioc2FwcGx5KHNhbXBsZV9uYW1lcywgZnVuY3Rpb24oeCl7c3Vic3RyKHgsIHN0YXJ0ID0gMywgc3RvcCA9IDQpfSkpCiAgICB0aXNzdWUgPC0gZmFjdG9yKHNhcHBseShzYW1wbGVfbmFtZXMsIGZ1bmN0aW9uKHgpe3N1YnN0cih4LCBzdGFydCA9IDUsIHN0b3AgPSA2KX0pKQogICAgY29uZGl0aW9uX2lkIDwtIGZhY3RvcihzYXBwbHkoc2FtcGxlX25hbWVzLCBmdW5jdGlvbih4KXtpZiAoZ3JlcGwoImtvIiwgeCwgZml4ZWQgPSBUUlVFKSkgezF9IGVsc2UgezB9fSkpCgogICAgIyBGb3IgREVzZXEyIHdlIG5lZWQgdG8gdHVybiB0aGlzIGludG8gYSBkYXRhZnJhbWUgCiAgICBzYW1wbGVfZGYgPSBkYXRhLmZyYW1lKHNhbXBsZV9uYW1lcyA9IHNhbXBsZV9uYW1lcywgY29uZGl0aW9uID0gY29uZGl0aW9uLCB0aW1lID0gdGltZSwgdGlzc3VlID0gdGlzc3VlLCBjb25kaXRpb25faWQ9Y29uZGl0aW9uX2lkKQogICAgZGRzX21hdCA8LSBERVNlcURhdGFTZXRGcm9tTWF0cml4KGNvdW50RGF0YSA9IGNvdW50X21hdHJpeCwKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgIGNvbERhdGEgPSBzYW1wbGVfZGYsCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICBkZXNpZ24gPSB+dGltZStjb25kaXRpb25faWQpICMgSGF2ZSB0aXNzdWUgYXMgYSBmYWN0b3IKICAgIAogICAgZGRzIDwtIGVzdGltYXRlU2l6ZUZhY3RvcnMoZGRzX21hdCkKICAgIAogICAgZGRzX21hdCRkZXNpZ24gIyBQcmludCB0aGUgZGVzaWduIG9mIHRoZSBleHBlcmltZW50CiAgICBkZHMgPC0gZXN0aW1hdGVTaXplRmFjdG9ycyhkZHNfbWF0KQogICAgCiAgICBudW1fc2FtcGxlc19tZWV0aW5nX2NyaXRlcmlhIDwtIDYgIyBiZSBzdHJpY3QgYW5kIGVuZm9yY2UgdGhhdCBhdCBsZWFzdCBoYWxmIHRoZSBzYW1wbGVzIG5lZWQgdG8gbWVldCB0aGUgY3JpdGVyaWEgKGkuZS4gb25lIGZ1bGwgY29uZGl0aW9uKQogICAgbnVtX2NvdW50c19pbl9nZW5lIDwtIDEwICAjIFRoZXkgbmVlZCBhdCBsZWFzdCAxMCBjb3VudHMKICAgIGtlZXAgPC0gcm93U3Vtcyhjb3VudHMoZGRzX21hdCkgPj0gbnVtX2NvdW50c19pbl9nZW5lKSA+PSBudW1fc2FtcGxlc19tZWV0aW5nX2NyaXRlcmlhCiAgICBkZHMgPC0gZGRzX21hdFtrZWVwLF0gIyBPbmx5IGtlZXAgdGhlIHJvd3Mgd2l0aCB0aGlzIGNyaXRlcmlhCiAgICAKICAgICMgTGV0J3MgcHJpbnQgdGhlIG51bWJlciBvZiByb3dzICYgY29sdW1uCiAgICBwcmludChwYXN0ZSgiRGF0YXNldCBkaW1lbnNpb25zOiAiLCBucm93KGRkcyksIG5jb2woZGRzKSkpCiAgICBpZiAobmNvbChkZHMpICE9IDEyKSB7CiAgICAgIHByaW50KHBhc3RlKCI9PT09PT09PT09PT09PT09PT0gV0FSTklORyBXQVJOSU5HIFdBUk5JTkcgWU9VUiBDT0xVTU5TIE1BWSBIQVZFIFRIRSBXUk9ORyBESU1TICA9PT09PT09PT09PT09PT09PT09PT09PT09PT0iLCBzZXA9IiIpKQogICAgfQogICAgCiAgICAjIFJ1biBERXNlcTIKICAgIGRkcyA8LSBERVNlcShkZHMpCiAgICAKICAgICMgQnVpbGQgcmVzdWx0cyB0YWJsZQogICAgcmVzIDwtIHJlc3VsdHMoZGRzKQogICAgcHJpbnQocGFzdGUoIkRlc2VxMiBkZXNpZ246ICIsIGRlc2lnbihkZHMpKSkKCiAgICAjIFN1bWFyaXNlIHRoZSByZXN1bHRzCiAgICBzdW1tYXJ5KHJlcykKICAgIAogICAgIyBMYXN0bHksIHdlIG1heSB3YW50IHRvIHNlZSB0aGUgcmVzdWx0cyBvZiB0aGUgaGlnaCBsb2dmb2xkY2hhbmdlIGUuZy4gPiAxIHdpdGggYSBwYWRqIHZhbHVlIDwgMC4wNQogICAgcmVzX3BhZGowNV9sZmMxIDwtIHJlc3VsdHMoZGRzLCBsZmNUaHJlc2hvbGQ9MikKICAgIHRhYmxlKHJlc19wYWRqMDVfbGZjMSRwYWRqIDwgMC4wNSkKICAgIAogICAgIyBTYXZlIHRoZSByZXN1bHRzCiAgICByZXNfb3JkZXJlZCA8LSByZXNbb3JkZXIocmVzJHB2YWx1ZSksXQogICAgb3V0cHV0X2ZpbGVuYW1lIDwtIHBhc3RlKHByb2plY3RfZGlyLCBvdXRwdXRfZmlsZW5hbWUsIHNlcD0nJykKICAgIHdyaXRlLmNzdihyZXNfb3JkZXJlZCwgZmlsZSA9IG91dHB1dF9maWxlbmFtZSkKICAgIHJldHVybihyZXNfb3JkZXJlZCkKfQoKYGBgCgojIyBSdW4gREUgb24gZWFjaCB0aXNzdWUKCkhlcmUgd2UgcnVuIHRoZSBERSBhbmFseXNpcyBvbiBlYWNoIHRpc3N1ZSBmb3IgdGhlIGNvbXBhcmlzb24uCmBgYHtyfQoKcnVuRGVzZXEyQmV0d2VlbkNvbmQocGFzdGUoJ21lcmdlZF9kZl9mYl9GRUFUVVJFX0NPVU5UU18nLCBkYXRlLCAnLmNzdicsIHNlcD0nJyksIHBhc3RlKCdERXNlcTJfQ05TX2ZiXycsIGRhdGUsICcuY3N2Jywgc2VwPScnKSkKcnVuRGVzZXEyQmV0d2VlbkNvbmQocGFzdGUoJ21lcmdlZF9kZl9tYl9GRUFUVVJFX0NPVU5UU18nLCBkYXRlLCAnLmNzdicsIHNlcD0nJyksIHBhc3RlKCdERXNlcTJfQ05TX21iXycsIGRhdGUsICcuY3N2Jywgc2VwPScnKSkKcnVuRGVzZXEyQmV0d2VlbkNvbmQocGFzdGUoJ21lcmdlZF9kZl9oYl9GRUFUVVJFX0NPVU5UU18nLCBkYXRlLCAnLmNzdicsIHNlcD0nJyksIHBhc3RlKCdERXNlcTJfQ05TX2hiXycsIGRhdGUsICcuY3N2Jywgc2VwPScnKSkKcnVuRGVzZXEyQmV0d2VlbkNvbmQocGFzdGUoJ21lcmdlZF9kZl9zY19GRUFUVVJFX0NPVU5UU18nLCBkYXRlLCAnLmNzdicsIHNlcD0nJyksIHBhc3RlKCdERXNlcTJfQ05TX3NjXycsIGRhdGUsICcuY3N2Jywgc2VwPScnKSkKCmBgYAoKIyMgTm9ybWFsaXNlIHRoZSBkYXRhCmBgYHtyfQpsaWJyYXJ5KGVkZ2VSKQoKZmlsZW5hbWUgPC0gcGFzdGUocHJvamVjdF9kaXIsICdtZXJnZWRfZGZfRkVBVFVSRV9DT1VOVFNfJywgZGF0ZSwgJy5jc3YnLCBzZXA9JycpCiMgaHR0cHM6Ly9iaW9jb25kdWN0b3Iub3JnL3BhY2thZ2VzL3JlbGVhc2Uvd29ya2Zsb3dzL3ZpZ25ldHRlcy9ybmFzZXFHZW5lL2luc3QvZG9jL3JuYXNlcUdlbmUuaHRtbApjb3VudHMgPC0gcmVhZC5jc3YoZmlsZW5hbWUsIGhlYWRlciA9IFRSVUUsIHNlcCA9ICIsIikKcm93bmFtZXMoY291bnRzKSA8LSBjb3VudHMkZW50cmV6Z2VuZV9pZApnZW5lcyA8LSBjb3VudHMkZW50cmV6Z2VuZV9pZAoKIyBMZXQncyBtYWtlIHN1cmUgb3VyIGNvdW50IGRhdGEgaXMgaW4gbWF0cml4IGZvcm1hdCBhbmQgaXMgb25seSB0aGUgbnVtZXJpYyBjb2x1bW5zIGkuZS4gZXZlcnl0aGluZyBidXQgdGhlIGdlbmVzCmNvdW50cyA8LSBjb3VudHNbLDI6bmNvbChjb3VudHMpXQpjb3VudHNbaXMubmEoY291bnRzKV0gPSAwCiMgSGVyZSB3ZSBnZXQgdGhlIG5hbWVzIG9mIHRoZSBjb2x1bW5zLCBmb3IgbXkgc3R1ZmYgSSBhbHdheXMgaGF2ZSBhbGwgdGhlIGluZm8gaW4gb25lIHN0cmluZyBhcyBJIGZpbmQgaXQgbWFrZXMgaXQgZWFzaWVyCiMgdGhpcyBtZWFucyBlYWNoIG9mIHRoZSAiZ3JvdXBzIiBhcmUgc2VwYXJhdGVkIGJ5IGEgIl8iIHRoaXMgbWF5IGJlIGRpZmZlcmVudCBmb3IgeW91CnNhbXBsZU5hbWVzIDwtIGNvbG5hbWVzKGNvdW50cykgIyBTYW1wbGUgbmFtZXMKCiMgRm9yIERFc2VxMiB3ZSBuZWVkIHRvIHR1cm4gdGhpcyBpbnRvIGEgZGF0YWZyYW1lIApzYW1wbGVERiA9IGRhdGEuZnJhbWUoc2FtcGxlTmFtZXMgPSBzYW1wbGVOYW1lcykKCiMgTWFrZSBzdXJlIHdlIGRvbid0IGluY2x1ZGUgdGhlIElEIGluIG91ciBjb2x1bW5zCmNvdW50TWF0cml4IDwtIGFzLm1hdHJpeChjb3VudHMpCgojIFdlIG5vdyBzZXQgdGhlIHJvdyBuYW1lcyB0byBiZSB0aGUgZ2VuZSBJRHMKcm93bmFtZXMoY291bnRNYXRyaXgpIDwtIGdlbmVzCgpkZHNNYXQgPC0gREVTZXFEYXRhU2V0RnJvbU1hdHJpeChjb3VudERhdGEgPSBjb3VudE1hdHJpeCwKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgY29sRGF0YSA9IHNhbXBsZURGLAogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICBkZXNpZ24gPSB+IDEpCgpkZHMgPC0gZXN0aW1hdGVTaXplRmFjdG9ycyhkZHNNYXQpCm91dHB1dEZpbGVuYW1lIDwtIHBhc3RlKHByb2plY3RfZGlyLCAnbWVyZ2VkX2RmX0ZFQVRVUkVfQ09VTlRTX0RFc2VxMk5vcm1fJywgZGF0ZSwgJy5jc3YnLCBzZXA9JycpCndyaXRlLmNzdihjb3VudHMoZGRzLG5vcm1hbGl6ZWQ9VFJVRSksIGZpbGUgPSBvdXRwdXRGaWxlbmFtZSkKCnJsZCA8LSBybG9nKGNvdW50TWF0cml4ICsgMSwgYmxpbmQgPSBGQUxTRSkKcm93bmFtZXMocmxkKSA8LSBnZW5lcwpvdXRwdXRGaWxlbmFtZSA8LSBwYXN0ZShwcm9qZWN0X2RpciwgJ21lcmdlZF9kZl9GRUFUVVJFX0NPVU5UU19ybG9nXycsIGRhdGUsICcuY3N2Jywgc2VwPScnKQp3cml0ZS5jc3YocmxkLCBmaWxlID0gb3V0cHV0RmlsZW5hbWUpCgojIEJhc2VkIG9uIHRoZSBkb2N1bWVudGF0aW9uIGJlbG93LCBpdCBpcyBiZXN0IHRvIHVzZSBUTU0gbm9ybWFsaXNhdGlvbiBmcm9tIEVkZ2VSIHdoZW4gCiMgY29tcGFyaW5nIGFjY3Jvc3MgZ2VuZXMsIHdoaWNoIGlzIHdoYXQgd2Ugd2lsbCBiZSBkb2luZyBpbiBvdXIgc3Vic2VxdWVudCBhbmFseXNlcy4KIyBodHRwczovL2hiY3RyYWluaW5nLmdpdGh1Yi5pby9ER0Vfd29ya3Nob3AvbGVzc29ucy8wMl9ER0VfY291bnRfbm9ybWFsaXphdGlvbi5odG1sCiMgaHR0cHM6Ly9yZHJyLmlvL2Jpb2MvZWRnZVIvbWFuL2NhbGNOb3JtRmFjdG9ycy5odG1sCiMgaHR0cHM6Ly93d3cuYmlvY29uZHVjdG9yLm9yZy9wYWNrYWdlcy9yZWxlYXNlL2Jpb2MvdmlnbmV0dGVzL2VkZ2VSL2luc3QvZG9jL2VkZ2VSVXNlcnNHdWlkZS5wZGYKIyBodHRwczovL3d3dy5iaW9zdGFycy5vcmcvcC8zMTc3MDEvCgpkZ2UgPC0gREdFTGlzdChjb3VudHM9Y291bnRNYXRyaXgpCmNwbShkZ2UpCmRnZSA8LSBjYWxjTm9ybUZhY3RvcnMoZGdlLCBtZXRob2Q9IlRNTSIpCnRtbSA8LSBjcG0oZGdlKQpyb3duYW1lcyh0bW0pIDwtIGdlbmVzCgpvdXRwdXRGaWxlbmFtZSA8LSBwYXN0ZShwcm9qZWN0X2RpciwgJ21lcmdlZF9kZl9GRUFUVVJFX0NPVU5UU190bW1fJywgZGF0ZSwgJy5jc3YnLCBzZXA9JycpCndyaXRlLmNzdih0bW0sIGZpbGUgPSBvdXRwdXRGaWxlbmFtZSwgcm93Lm5hbWVzPVRSVUUpCgp2c2QgPC0gdnN0KGRkcywgYmxpbmQgPSBUUlVFKQpyb3duYW1lcyh2c2QpIDwtIGdlbmVzCm91dHB1dEZpbGVuYW1lIDwtIHBhc3RlKHByb2plY3RfZGlyLCAnbWVyZ2VkX2RmX0ZFQVRVUkVfQ09VTlRTX3ZzdF8nLCBkYXRlLCAnLmNzdicsIHNlcD0nJykKd3JpdGUuY3N2KHJsZCwgZmlsZSA9IG91dHB1dEZpbGVuYW1lKQoKYGBgCgojIyMgUHJpbnQgc2Vzc2lvbiBpbmZvCmBgYHtyfQpzZXNzaW9uSW5mbygpCmBgYAoKIyMjIFJlZmVyZW5jZXM6CmhiY3RyYWluaW5nL0RHRV93b3Jrc2hvcC4gVGVhY2hpbmcgbWF0ZXJpYWxzIGF0IHRoZSBIYXJ2YXJkIENoYW4gQmlvaW5mb3JtYXRpY3MgQ29yZSwgMjAyMS4gQWNjZXNzZWQ6IEp1bi4gMTUsIDIwMjEuIFtPbmxpbmVdLiBBdmFpbGFibGU6IGh0dHBzOi8vZ2l0aHViLmNvbS9oYmN0cmFpbmluZy9ER0Vfd29ya3Nob3AgIAoKTG92ZSBNSSwgSHViZXIgVywgQW5kZXJzIFMgKDIwMTQpLiDigJxNb2RlcmF0ZWQgZXN0aW1hdGlvbiBvZiBmb2xkIGNoYW5nZSBhbmQgZGlzcGVyc2lvbiBmb3IgUk5BLXNlcSBkYXRhIHdpdGggREVTZXEyLuKAnSBHZW5vbWUgQmlvbG9neSwgMTUsIDU1MC4gZG9pOiAxMC4xMTg2L3MxMzA1OS0wMTQtMDU1MC04LiAgCgpNLiBELiBSb2JpbnNvbiwgRC4gSi4gTWNDYXJ0aHksIGFuZCBHLiBLLiBTbXl0aCwg4oCYZWRnZVI6IGEgQmlvY29uZHVjdG9yIHBhY2thZ2UgZm9yIGRpZmZlcmVudGlhbCBleHByZXNzaW9uIGFuYWx5c2lzIG9mIGRpZ2l0YWwgZ2VuZSBleHByZXNzaW9uIGRhdGHigJksIEJpb2luZm9ybWF0aWNzLCB2b2wuIDI2LCBuby4gMSwgcHAuIDEzOeKAkzE0MCwgSmFuLiAyMDEwLCBkb2k6IDEwLjEwOTMvYmlvaW5mb3JtYXRpY3MvYnRwNjE2LiAgCgo=