A networks analysis pipeline for RNASeq time series data.
NetSeekR is a network analysis R package that includes the capacity to analyze time series of RNASeq data, perform correlation and regulatory network inferences and use network analysis methods to summarize the results of a comparative genomics study.
Authors: Himangi Srivastava, Drew Ferrell, and George V. Popescu.
The NetSeekR code requires specific versions for packages that are used.
| Package | Version |
|---|---|
| pacman | 0.5.1 |
| BiocManager | 1.30.10 |
| magrittr | 1.5 |
| readr | 1.3.1 |
| purrr | 0.4.2 |
| stringr | 0.3.3 |
| ggplot2 | 1.4.0 |
| devtools | 3.2.1 |
| flashClust | 2.2.1 |
| tidyr | 1.01-2 |
| networkD3 | 1.0.0 |
| igraph | 0.4 |
| limma | 1.2.4.2 |
| edgeR | 3.42.0 |
| topGO | 2.37.0 |
| WGCNA | 1.68 |
| biomaRt | 2.42.0 |
| Rgraphviz | 2.30.0 |
| dplyr | 0.8.3 |
Below are the steps to run NetSeekR.
- Set the working directory to the NetSeekR path.
setwd(<<path/to/NetSeekR>>)
- Unzip the NetSeekR file.
unzip('NetSeekR.zip')
- Load packages and source functions for NetSeekR.
source('scripts/NetSeekR.R')
- Edit configuration file and sample comparison matrix.
- note Below is a template configuration file which needs to be edited per usage.
| analysis_type | custom tag |
|---|---|
| design_matrix | path to experimental design matrix |
| edger_adjustment_method | edgeR: p-value adjustment method |
| edger_lfc | limma: minimum log2-fold-change that is considered scientifically meaningful |
| edger_method | NOT USED |
| feature_counts_path | path to feature counts software |
| kallisto | Boolean value for Kallisto execution decision |
| kallisto_bias | sequence based bias correction |
| kallisto_bootstrap_samples | bootstrap sample number |
| kallisto_chromosomes | tab separated file with chromosome names and lengths |
| kallisto_fasta_files | path to genome annotation file |
| kallisto_fastq_files | reads to quantify |
| kallisto_fr_stranded | strand specific reads, first read forward |
| kallisto_fragment_length | estimated average fragment length |
| kallisto_fusion | search for fusions for Pizzly |
| kallisto_genomebam | project pseudoalignments to genome sorted BAM file |
| kallisto_gtf | GTF file for transcriptome information (required for --genomebam) |
| kallisto_index | location to write genome index from Kallisto (required for Kallisto alignment) |
| kallisto_kmer_size | k-mer (odd) length (defaut: 31, max value: 31 |
| kallisto_make_unique | replace repeated target names with unique names |
| kallisto_output_dir | directory to write quantification output to |
| kallisto_path | path to Kallisto software |
| kallisto_plaintext | output plaintext instead of HDF5 |
| kallisto_pseudobam | save pseudoalignments to transcriptome to BAM file |
| kallisto_rf_stranded | strand specific reads, first read reverse |
| kallisto_sd | estimated standard deviation of fragment length (default: -l, -s values are estimated from paired end data, but are required when using --single) |
| kallisto_seed | seed for the bootstrap sampling (default: 42) |
| kallisto_single | quantify single-end reads |
| kallisto_single_overhang | include reads where unobserved rest of fragment is predicted to lie outside a transcript |
| kallisto_threads | number of threads to use (default: 1) |
| sample_comparisons_file | path to the sample comparison file for differential gene expression testing |
| sample_covariates | experimental design matrix column names to be used as covariates with Sleuth |
| significance_cutoff | a cutoff value for determining significance |
| sleuth_gene_mode | Boolean value for Sleuth gene mode execution decision |
| sleuth_transcript_mode | Boolean value for Sleuth transcript mode execution decision |
| star | Boolean value for STAR execution decision |
| star_genomeDir | path to the directory where the genome indices are stored |
| star_genomeFastaFiles | path to a FASTA file with the genome reference sequences |
| star_path | path to STAR software |
| star_readFilesIn | path to the folder containing the sequences to be mapped |
| star_runThreadN | number of threads to be used for genome generation, it has to be set to the number of available cores on the server node |
| star_sjdbGTFfile | path to the file with annotated transcripts in the standard GTF format |
| star_sjdbOverhang | length of the genomic sequence around the annotated junction to be used in constructing the splice junctions database> |
- Align batches of reads.
alignment_results <- implement_alignment(arguments_file = <<path/to/configuration file>>)
- Test for differential gene expression.
- note An edited sample comparison matrix needs to be supplied for differential gene expression testing. An example of an edited sample comparison matrix file is below. A sample comparison matrix does not need headers, only sample identifiers. Which samples to compare in differential testing should be written row-wise.
| Example | |||||
|---|---|---|---|---|---|
| SL209924 | SL209925 | SL209926 | SL209921 | SL209922 | SL209923 |
| SL209927 | SL209928 | SL209937 | SL209921 | SL209922 | SL209923 |
| SL209938 | SL209939 | SL209940 | SL209921 | SL209922 | SL209923 |
| SL209944 | SL209945 | SL209946 | SL209941 | SL209942 | SL209943 |
| SL209947 | SL209948 | SL209949 | SL209941 | SL209942 | SL209943 |
| SL209950 | SL209951 | SL209952 | SL209941 | SL209942 | SL209943 |
implement_differential_gene_expression(alignment_results)
- Network analysis is then conducted assuming sets of differentially expressed genes are available.
implement_network_analysis(alignment_tool = 'star', alignment_results = alignment_results)
The network analysis function above will produce an image of the network. The image can be saved from the ‘Export’ tab at the top of the window.