ATpoint

get_rle_ranges <- function(x) {
  rle_out <- rle(x)
  rle_lengths <- rle_out$lengths
  ends <- cumsum(rle_lengths)
  starts <- c(1, head(ends, -1) + 1)
  lens <- (ends - starts) + 1
  data.frame(run = seq_along(rle_lengths), start = starts, end = ends, length = lens)
}

ATpoint

Some reminders useful when admining a server

# List all files with that extension 1GB or larger

find . -type f -size +1G -print | grep -E "\\.lif$|\\.lifext$" > lif_files.txt

# Print combined size of all files in that file in GB
awk '
 {

ATpoint

# Say we have a SingleCellExperiment "se"
nf <- se$norm.factors
ls <- se$lib.size

# effective library size
els <- ls * nf

# size factors centered around 1
geomeans <- exp(mean(log(els)))
sf <- els / geomeans

ATpoint

library(tidyverse)
library(ReactomePA)
library(biomaRt)

r_human <- ReactomePA:::get_Reactome_DATA("human")

# This package maps to Entrez so we need the conversion table to Ensembl IDs
human_entrez2ensembl <- getBM(
  attributes = c("entrezgene_id", "ensembl_gene_id"),
  mart = mart_human

ATpoint

# Retrieve GO terms in R using AnnotationDbi and org.Mm.eg.db

library(org.Mm.eg.db)
library(AnnotationDbi)

all_go <- AnnotationDbi::select(
  org.Mm.eg.db,
  keys = keys(org.Mm.eg.db, keytype = "ENTREZID"),
  columns = c("GO", "ONTOLOGY", "ENSEMBL"),
  keytype = "ENTREZID"

ATpoint

pkgs <- c("KEGGREST", "org.Mm.eg.db", "tidyverse", "AnnotationDbi")
invisible(lapply(pkgs, function(x) suppressPackageStartupMessages(library(x, character.only = TRUE))))

pathway_id_2_name <- 
  keggList("pathway", "mmu") %>%
  enframe(name = "pathway_id", value = "pathway_name") %>%
  mutate(pathway_id = gsub("mmu", "", pathway_id),
         pathway_name = gsub(" - .*", "", pathway_name))

pathway_2_entrez <- 

ATpoint

#!/bin/bash

# Read existing md5sums and use to confirm integrity of fastq files

#SBATCH --nodes=1
#SBATCH --cpus-per-task=36
#SBATCH --partition=normal
#SBATCH --time=08:00:00
#SBATCH [email protected]
#SBATCH --job-name=md5checker

ATpoint

# Install R infrastructure
pkg_install <- c("reticulate", "zellkonverter")
BiocManager::install(pkg_install, update = FALSE)

# Install conda itself via the reticulate package, takes time...
reticulate::install_miniconda()

# Create an environment for velocity with a specific python version that is required
# to install a downgraded matplotlib, see below
reticulate::conda_create(envname = "velocity", python_version = "3.8.19")

ATpoint

geom_segment(data=metadata(sce)$grid.df, 
    mapping=aes(x=start.1, y=start.2, xend=end.1, yend=end.2), 
    size=1,
    arrow=arrow(length=unit(0.1, "inches"), type="closed"), inherit.aes = FALSE)

ATpoint

#' Read loom files into R as CsparseMatrix, depending on Matrix and hdf5r
#' @param path to file loom file on disk
#' @param include.ambiguous logical, whether to also read the "ambiguous" counts from velocyto, default FALSE
#'
readLoomMatrices <- function(file, include.ambiguous=FALSE) {
  
  require(Matrix)
  require(hdf5r)
  
  engine='hdf5r'