init

latch-verified · Aug 2, 2022 · 1c2115a · 1c2115a
commit 1c2115a
Show file tree

Hide file tree

Showing 16 changed files with 10,867 additions and 0 deletions.
diff --git a/.envrc b/.envrc
@@ -0,0 +1,3 @@
+#!/usr/bin/env bash
+eval "$(conda shell.bash hook)"
+conda activate wf-core-go-pathway
diff --git a/.gitignore b/.gitignore
@@ -0,0 +1,7 @@
+__TESTING__/
+.latch_report.tar.gz
+wf/__pycache__/
+build/
+node_modules/
+
+
diff --git a/Dockerfile b/Dockerfile
@@ -0,0 +1,104 @@
+FROM 812206152185.dkr.ecr.us-west-2.amazonaws.com/latch-base:02ab-main
+
+SHELL ["/usr/bin/env", "bash", "-c"]
+
+
+# Allow --mount=cache to do its job
+# https://github.com/moby/buildkit/blob/86c33b66e176a6fc74b88d6f46798d3ec18e2e73/frontend/dockerfile/docs/syntax.md#run---mounttypecache
+RUN rm /etc/apt/apt.conf.d/docker-clean
+RUN echo 'Binary::apt::APT::Keep-Downloaded-Packages "true";' > /etc/apt/apt.conf.d/keep-cache
+
+
+# Generic installation dependencies
+# wget - obvious
+# software-properties-common - `add-apt-repository`
+# dirmngr - GPG key manager, loads from `/etc/apt/trusted.gpg.d/`
+RUN apt-get update && apt-get install --yes --no-install-recommends \
+  wget \
+  software-properties-common \
+  dirmngr
+
+
+#
+# R
+#
+
+# >>> Install R
+# https://cloud.r-project.org/bin/linux/debian/
+# https://github.com/rocker-org/rocker-versioned2/blob/f3325b2cf88d8899ddcb2f0945aa9f87ad150cd7/scripts/install_R_ppa.sh
+RUN apt-key adv --keyserver keyserver.ubuntu.com --recv-key '95C0FAF38DB3CCAD0C080A7BDC78B2DDEABC47B7'
+RUN add-apt-repository "deb https://cloud.r-project.org/bin/linux/debian $(lsb_release --codename --short)-cran40/"
+
+RUN apt-get update && apt-get install --yes \
+    r-base \
+    r-base-dev \
+    locales
+
+RUN apt-mark hold r-base r-base-dev
+
+RUN echo "en_US.UTF-8 UTF-8" >> /etc/locale.gen
+RUN locale-gen en_US.utf8
+RUN /usr/sbin/update-locale LANG="en_US.UTF-8"
+
+# >>> R packages
+RUN apt-get update
+RUN apt install --yes \
+    libcurl4-openssl-dev \
+    libxml2-dev \
+    libssl-dev
+
+RUN Rscript -e 'install.packages("BiocManager")'
+RUN Rscript -e 'BiocManager::install(version = "3.15")'
+RUN Rscript -e 'BiocManager::install(c( \
+  "purrr", \
+  "dplyr", \
+  "tibble", \
+  "readr", \
+  "readxl", \
+  "stringr", \
+  "vctrs", \
+  "clusterProfiler", \
+  "DOSE", \
+  "ggridges", \
+  "enrichplot", \
+  "ggplot2", \
+  "msigdbr", \
+  "pathview" \
+  ), update=FALSE)'
+
+RUN Rscript -e 'BiocManager::install(c( \
+  "org.Hs.eg.db", \
+  "org.Mm.eg.db", \
+  "org.Rn.eg.db", \
+  "org.Dm.eg.db", \
+  "org.At.tair.db", \
+  "org.Sc.sgd.db", \
+  "org.Dr.eg.db", \
+  "org.Ce.eg.db", \
+  "org.Bt.eg.db", \
+  "org.Ss.eg.db", \
+  "org.Gg.eg.db", \
+  "org.Mmu.eg.db", \
+  "org.Cf.eg.db", \
+  "org.EcK12.eg.db", \
+  "org.Xl.eg.db", \
+  "org.Ag.eg.db", \
+  "org.Pt.eg.db", \
+  "org.EcSakai.eg.db", \
+  "org.Mxanthus.db" \
+  ), update=FALSE)'
+
+RUN python3 -m pip install --upgrade latch imagesize jinja2
+
+# todo: automatically build frontend in Dockerfile
+
+# STOP HERE:
+# The following lines are needed to ensure your build environement works
+# correctly with latch.
+COPY ./report/build/index.html /root/template.html
+COPY ./go_pathway.r /root/go_pathway.r
+COPY wf /root/wf
+
+ARG tag
+ENV FLYTE_INTERNAL_IMAGE $tag
+WORKDIR /root
diff --git a/README.md b/README.md
@@ -0,0 +1,42 @@
+<html>
+<p align="center">
+  <img src="https://user-images.githubusercontent.com/31255434/182289305-4cc620e3-86ae-480f-9b61-6ca83283caa5.jpg" alt="Latch Verified" width="100">
+</p>
+
+<h1 align="center">
+  Pathway Analysis
+</h1>
+
+<p align="center">
+<strong>
+Latch Verified
+</strong>
+</p>
+
+<p align="center">
+  Implicate pathways + gene ontologies from a set of desired genes.
+</p>
+
+<p align="center">
+  <a href="https://github.com/latch-verified/pathway/releases/latest">
+    <img src="https://img.shields.io/github/release/latch-verified/pathway.svg" alt="Current Release" />
+  </a>
+  <a href="https://github.com/latch-verified/pathway/actions/workflows/tests.yml">
+    <img src="https://github.com/latch-verified/pathway/actions/workflows/tests.yml/badge.svg" alt="End-to-End Tests" />
+  </a>
+  <a href="https://opensource.org/licenses/MIT">
+    <img src="https://img.shields.io/badge/LICENSE-MIT-brightgreen.svg" alt="License" />
+  </a>
+  <img src="https://img.shields.io/github/commit-activity/w/latch-verified/pathway.svg?style=plastic" alt="Commit Activity" />
+  <img src="https://img.shields.io/github/commits-since/latch-verified/pathway/latest.svg?style=plastic" alt="Commits since Last Release" />
+</p>
+
+<h3 align="center">
+  <a href="https://console.latch.bio/explore/65992/info">Hosted Interface</a>
+  <span> · </span>
+  <a href="https://docs.latch.bio">SDK Documentation</a>
+  <span> · </span>
+  <a href="https://join.slack.com/t/latchbiosdk/shared_invite/zt-193ibmedi-WB6mBu2GJ2WejUHhxMOuwg">Slack Community</a>
+</h3>
+
+</html>
diff --git a/go_pathway.r b/go_pathway.r
@@ -0,0 +1,159 @@
+options(show.error.locations = TRUE)
+
+p <- function (...) {
+  cat(as.character(...), sep = "\n")
+}
+
+warn <- function(...) {
+  output <- paste(
+    "__LATCH_WARNING_START__",
+    as.character(...),
+    "__LATCH_WARNING_END__",
+    sep = "\n"
+  )
+  cat(output)
+}
+
+p("Importing")
+suppressMessages(suppressWarnings(library(vctrs)))
+suppressMessages(suppressWarnings(library(dplyr)))
+suppressMessages(suppressWarnings(library(tibble)))
+suppressMessages(suppressWarnings(library(readr)))
+suppressMessages(suppressWarnings(library(readxl)))
+suppressMessages(suppressWarnings(library(stringr)))
+
+suppressMessages(suppressWarnings(library(purrr)))
+
+suppressMessages(suppressWarnings(library(clusterProfiler)))
+suppressMessages(suppressWarnings(library(DOSE)))
+suppressMessages(suppressWarnings(library(ggridges)))
+suppressMessages(suppressWarnings(library(enrichplot)))
+suppressMessages(suppressWarnings(library(ggplot2)))
+suppressMessages(suppressWarnings(library(msigdbr)))
+suppressMessages(suppressWarnings(library(pathview)))
+
+read_tabular = function (path) {
+  tryCatch({
+    read_excel(path)
+  }, error = function (cond) {
+    read_delim(path, trim_ws = TRUE)
+  })
+}
+
+p("Loading gene annotations")
+organism = "org.Hs.eg.db"
+suppressMessages(suppressWarnings(library(organism, character.only=TRUE)))
+msig_db <- msigdbr(species="Homo sapiens") %>% dplyr::select(gs_name, entrez_gene)
+
+p("Loading contrast data")
+args <- commandArgs(trailingOnly = TRUE)
+
+res <- read_tabular(args[[1]])
+num_pathways <- strtoi(args[[2]])
+
+p("Preparing data")
+dsD <- res %>%
+  mutate(gene_name = .data[[vec_as_names("", repair = "unique")]]) %>%
+  dplyr::select(gene_name, log2FoldChange) %>%
+  na.omit %>%
+  mutate(gene_name = mapIds(org.Hs.eg.db, keys=gene_name, keytype="ALIAS", column="ENTREZID")) %>%
+  na.omit %>%
+  distinct(gene_name, .keep_all=TRUE) %>%
+  column_to_rownames("gene_name")
+ds <- dsD$log2FoldChange
+names(ds) <- rownames(dsD)
+ds <- ds %>% sort(decreasing=TRUE)
+
+p("Running MSig")
+start <- Sys.time()
+msig <- GSEA(ds, TERM2GENE=msig_db)
+print(Sys.time() - start)
+
+head(msig)
+p("Running GO")
+start <- Sys.time()
+go <- gseGO(ds, ont="ALL", organism, keyType="ENTREZID")
+print(Sys.time() - start)
+
+head(go)
+
+if (nrow(go) > 0) {
+  p("  Plotting")
+  dir.create("/root/res/Gene Ontology", showWarnings = FALSE, recursive = TRUE)
+  png(file="/root/res/Gene Ontology/Dot Plot.png", width=960, height=900)
+  print(dotplot(go, showCategory=20, split=".sign") + facet_grid(.~.sign))
+  dev.off()
+
+  png(file="/root/res/Gene Ontology/Ridge Plot.png", width=960, height=900)
+  ridgeplot(go) + labs(x = "enrichment distribution")
+  dev.off()
+} else {
+  warn(paste(
+    "No statistically significant enriched gene sets (with cutoff p ≤ 0.05) found after ",
+    "running gene set enrichment analysis (GSEA) on Gene Ontology.",
+    sep = ""
+  ))
+}
+
+p("Running KEGG")
+start <- Sys.time()
+kks <- gseKEGG(ds, "hsa")
+print(Sys.time() - start)
+
+
+if (nrow(kks) > 0) {
+  p("  Plotting")
+  dir.create("/root/res/KEGG", showWarnings = FALSE, recursive = TRUE)
+  png(file="/root/res/KEGG/Dot Plot.png", width=960, height=900)
+  print(dotplot(kks, showCategory=20, split=".sign") + facet_grid(.~.sign))
+  dev.off()
+
+  png(file="/root/res/KEGG/Ridge Plot.png", width=960, height=900)
+  ridgeplot(kks) + labs(x = "enrichment distribution")
+  dev.off()
+
+  entrezIDsToGeneNames <- function (entrezIDs) {
+    return(mapIds(org.Hs.eg.db, entrezIDs, "SYMBOL", "ENTREZID") %>% paste(collapse = " "))
+  }
+
+  pathways <- kks@result %>%
+    slice_max(order_by = enrichmentScore, n = num_pathways) %>%
+    mutate(coreEntrezIDs = core_enrichment) %>%
+    mutate(entrezList = strsplit(core_enrichment, '/'), .keep = "unused") %>%
+    mutate(coreEnrichedGenes = unlist(lapply(entrezList, entrezIDsToGeneNames)), .keep = "unused")
+
+  write.csv(pathways, "/root/res/KEGG/table.csv", row.names = FALSE)
+
+  dir.create("/root/tempres", showWarnings = FALSE, recursive = TRUE) 
+  genesets_path <- "/root/tempres/genesets.txt"
+  geneSets <- kks@geneSets
+  write("PATHWAYIDS", genesets_path)
+  lapply(names(geneSets), write, genesets_path, append=TRUE, ncolumns=100000)
+  write("ENTREZIDS", genesets_path, append=TRUE)
+  lapply(geneSets, write, genesets_path, append=TRUE, ncolumns=100000)
+  write("NAMES", genesets_path, append=TRUE)
+  geneNames <- lapply(geneSets, entrezIDsToGeneNames)
+  lapply(geneNames, write, genesets_path, append=TRUE, ncolumns=100000)
+
+  for (pathwayID in pathways$ID) {
+    p(paste("  Running pathview on", pathwayID))
+    pathview(gene.data=ds, pathway.id=pathwayID, species="Homo sapiens")
+  }
+} else {
+  warn(paste(
+    "No statistically significant enriched gene sets (with cutoff p ≤ 0.05) found after ",
+    "running gene set enrichment analysis (GSEA) on KEGG.",
+    sep = ""
+  ))
+}
+
+
+p("  Plotting")
+dir.create("/root/res/MSig", showWarnings = FALSE, recursive = TRUE)
+png(file="/root/res/MSig/Dot Plot.png", width=960, height=900)
+print(dotplot(msig, showCategory=20, split=".sign") + facet_grid(.~.sign))
+dev.off()
+
+png(file="/root/res/MSig/Ridge Plot.png", width=960, height=900)
+ridgeplot(msig) + labs(x = "enrichment distribution")
+dev.off()
diff --git a/report/.gitignore b/report/.gitignore
@@ -0,0 +1,23 @@
+# See https://help.github.com/articles/ignoring-files/ for more about ignoring files.
+
+# dependencies
+/node_modules
+/.pnp
+.pnp.js
+
+# testing
+/coverage
+
+# production
+/build
+
+# misc
+.DS_Store
+.env.local
+.env.development.local
+.env.test.local
+.env.production.local
+
+npm-debug.log*
+yarn-debug.log*
+yarn-error.log*