Repository navigation
Annotation module
The annotation module has the following main features:
-maps reads to reference genome;
-quantifies ncRNAs, miRNAs and IsomiRs
-tests for differential expression
-predicts novel miRNAs
/results_IsomiR_window
then inside this one create subdirectories
/bowtie_stats --> files with bowtie alignment statistics
/Log_processing_files --> files with alignment log
/Log_done_files --> log for rRNA filtering
/filtered_SAM --> in this directory results of the alignment file are produced. *filtered.sam: alignment file filtered for rRNA reads
/individual_SAM_results --> original alignment file before filtering for rRNA
/Htseq_outputs/individual_ncRNAs --> counts per sncRNA feature. miRNAs are not counted at this stage
/Htseq_outputs/counts_ncRNAs --> counts per snRNA category.
/fastq_file --> directory with fastq files (files should be trimmed for adapters and advised to be already filtered for read quality)
/Knowledge_bases --> directory with genome assemblies (fasta and indexed genomes), RNAcentral.gff3 and miRBase.gff and 3'UTR (to receive these knowledge bases please send email to isomirwindow1.0@gmail.com write the species code in subject. Then you will receive a link to transfer all the required files.
Species code
Bos_taurus
Canis_familiaris
Capra_hircus
Danio_rerio
Drosophila_melanogaster
Equus_caballus
Gallus_gallus
Homo_sapiens
Mus_musculus
Ovis_aries
Pan_troglodytes
Rattus_norvegicus
Sus_scrofa
Arabidopsis_thaliana
Brachypodium_distachyon
Oryza_sativa
Solanum_lycopersicum
Sorghum_bicolor
Vitis_vinifera
Zea_mays
It requires Species.pl and species_hash.txt
NM= number of mismatches for alignment
NMH= number of allowed multiple hits for read mapping
Bowtie runs in --best strata-- setting
running command
perl <directoryforscript>/find_ncRNAs.pl <pathto>/Knowledge_bases./ <pathto>/fastq_file <pathto>/results_IsomiR_window/bowtie_stats, <pathto>/Log_processing_files <pathto>/Log_done_files <pathto>/filtered_SAM <pathto>/individual_SAM_results <pathto>/Htseq_outputs/individual_ncRNAs <pathto>/Htseq_outputs/counts_ncRNAs <pathto>/fastq_file/<name_fastq_file> <NM> <NMH> <species code>
/find_isomirs_results /find_isomirs_results/isomirs /find_isomirs_results/isomirs/ambiguous /find_isomirs_results/isomirs/not_ambiguous /find_isomirs_results/SAMs_VCFs
Requires: Species.pl and species_hash.txt
running command
perl <directoryforscript>/find_isomiRs.pl <pathto>/Knowledge_bases <pathto>/filtered_SAM <pathto>/individual_SAM_results <pathto>/not_ambiguous <pathto>/ambiguous <pathto>/filtered_SAM/<nameof_filtered_SAMfile> <species code>
Step 6 - Download scripts: Deseq.pl and Rdeseq.R. These scripts prepare matrix of isomiR counts and of ncRNA counts across all replicates in two experimental conditions and perform testing for differential expression using Deseq.
this script has been created to generate read count matrices for ncRNAs and for IsomiRs of two experimental conditions so it requires the existence of the following directories:
C1/find_isomirs_results/isomirs/not_ambiguous C2/find_isomirs_results/isomirs/not_ambiguous /C1/Htseq_outputs/individual_ncRNAs /C2/Htseq_outputs/individual_ncRNAs /input_DESeq
running command
perl <directoryforscript>/deseq.pl <jobID> <pathto>/C1/find_isomirs_results/isomirs/not_ambiguous <pathto>/C2/find_isomirs_results/isomirs/not_ambiguous <pathto>/C1/Htseq_outputs/individual_ncRNAs <pathto>/C2/Htseq_outputs/individual_ncRNAs /input_DESeq <Unpaired_Sample or Paired_Sample>
after this step the required inputs for differential expression testing for IsomiRs and for ncRNAs (written at/input_DESeq) are prepared.
then create directory to send Deseq analysis output.
/output_DESeq
**running command**
Rscript Rdeseq.R <jobID> /input_DESeq /output_DESeq <choose design: paired or unpaired> <choose p-value: 0.05 or 0.10>
The outputs are: a) file CompleteTableDE_''.txt --> a Tab delimited matrix with the results of the Differential Expression Test order by p-value (smallest to biggest) and including the normalized counts of each replicate in each condition. b) file TableDE_''.txt' --> a Tab delimited matrix with the results of the Differential Expression Test, for IsomiRs with the p-value <= value set for the analysis including the normalized counts of each replicate in each condition. c) file MAplot_''.png --> MA plot d) file MAplot_''.pdf --> MA plot e) file heatmap_''.png --> heatmap displaying hierarquical clustering of samples including the top15 differentially expressed IsomiRs.
Step 7 - Download charts.pl and run it to generate summary stats of previous analysis: read length distribution, ncRNA counts, isomiR's Biogenesis, frequency of IsomiR types and miRNA normalized count matrix
this script allows to create summary tables that can be used to produce bar charts of read length distribution, piecharts of the proportion of ncRNAs and of miRNA biogenesis (3' or 5' arm) and it further creates a matrix of normalized counts at miRNA level aggregating the complexity of isomIRs.
Requires: Species.pl
requires creating a directory for results: /charts_C1 /charts_C2 /summary_tables
running command
perl charts.pl <jobID> <pathto>/C1/find_isomirs_results/isomirs/not_ambiguous <pathto>/C2/find_isomirs_results/isomirs/not_ambiguous <pathto>/C1/find_isomirs_results/isomirs/ambiguous <pathto>/C2/find_isomirs_results/isomirs/ambiguous
<pathto>/C1/Htseq_outputs/individual_ncRNAs <pathto>/C2/Htseq_outputs/individual_ncRNAs <pathto>/charts_C1 <pathto>/charts_C2 <pathto>/summary_tables <speciescode>
The outputs are:
Total_ncRNAs_piechart_C#.txt Types_isomiRs_C#.txt Counts_3p5p_C#.txt Counts_editings_C#.txt Counts_tailings_C#.txt Counts_families_C#.txt Final_TableDE_.txt --> summary statistics of isomiR counts for each condition (mean and stdev) miRNA_normalized_counts_.txt --> normalized miRNA counts for each replicate in each condition
Step 8 - Download script mirdeep.pl and merged_arf_and_collapsed_id.pl to perform miRNA prediction and incorporation of novel miRNAs in the species miRNA annotation
The IsomiR Window pipeline provides a script that can be used before step 5. This script enables de identification of novel miRNAs based in your data and allows if you desire you can add these novel miRNAs to the annotation of miRNAs of the corresponding species. In case you execute this script before find_isomiRs.pl, you will be able to identify the isomiRs that have been generated for these novel miRNAs as well. The script executes automatically miRdeep2 in caso of animal genomes or miRD2 in case of plant genomes. These predictions sometimes predict overlapping miRA percursors that overlap >=90%, the script creates an output that removes this redundancy, displaying in this case only the predictions with better score and better sequencing coverage. Nevertheless the full set of results is kept.
You will see that in the command flag options /filtered_SAM is repeated twice, this is due to the fact that the IsomiR Window web interface is designed for the comparison of two experimental conditions. In case you have that design you will use /C1/filtered_SAM and /C2/filtered_SAM, otherwise you just repeat the same path twice.
Requires: merged_arf_and_collapsed_id.pl
Before you run the script you must create two new folders:
/find_isomirs_results/miRprediction_input --> here it will be written the input files required for miRDeep2 (miRNA prediction for animal genomes) or miRDP2 (miRNA prediction for plant genomes) and that are created from the SAM file input
/find_isomirs_results/miRprediction_results--> here it will be written the results of the novel miRNA prediction
running command
perl <directoryforscript>/mirdeep.pl <jobID> <pathto>/Knowledge_bases <pathto>/filtered_SAM <pathto>/filtered_SAM /find_isomirs_results/miRprediction_input /find_isomirs_results/miRprediction_results yes <species code> <'C1', 'C2' ou 'both'> <minimum read stack (10, 50 ou 100)> <use annotation in prediction ("yes", "no")> <add novel miRNAs to annotation ("yes", "no")>