nf-core/stableexpression

Edit

Gene tables

For each section of genes, a table is displayed with the gene names and their statistics (including their rank, stability score).

Gene normalised expression

For each section, the normalised expression values are displayed of the top 25 genes are displayed as boxplots.

Dataset statistics

Various statistics about the analysed datasets are displayed, including the number of genes, the mean and standard deviation of the normalised expression values.

Skewness of normalised expression values:

Proportion of zero values:

Proportion of missing values (genes not present but present in other samples):

Dataset filtering

For each dataset, samples displaying a too high proportion of zero values or missing values are filtered out. The filtered datasets are then used for further analysis. The report displays the proportion of samples filtered out for each dataset. Users can modify the filtering thresholds using the --max_zero_ratio and --max_null_ratio parameters respectively.

Proportions of samples filtered out based on zero values:

Proportions of samples filtered out based on missing values:

Gene ID mapping

The pipeline maps gene IDs to their corresponding gene names using the Ensembl database.

Distribution of gene occurrences:

NB: The pipeline filters out genes based on gene ID occurrence using two parameters: --min_occurrence_freq and --min_occurrence_quantile. The first parameter filters out genes showing a too low frequency of occurrence, while the second parameter filters genes belonging to the last quantiles of occurrence. Users can tweak these parameters to increase or decrease the filtering stringency. The distribution of gene occurrences displayed in the MultiqcQC report can be very useful to detect a plateau of gene ID occurrence frequency, which can indicate the presence of rare gene IDs. In the distribution above, a plateau is reached around 0.236. In such case, setting --min_occurrence_freq 0.24 (default: 0.2) will remove all gene IDs with an occurrence frequency below this plateau.

Gene ID mapping statistics:

Reasons of failures of gene ID mapping:

Reasons of incomplete gene ID mapping:

Downloaded datasets

The report shows metadata about the downloaded datasets:

• Expression Atlas
• optionally: GEO datasets (when --fetch_geo_accessions was set)

The report also has a section for datasets that could not be downloaded, and the reasons for failure.

Example for Expression Atlas:

Software version

You can access versions of all softwares used in the pipeline, step by step:

To launch the app, you must first create and activate the appropriate conda environment:

conda env create -n dash_app -f reporting/dash_app/environment.yml
conda activate dash_app

then:

cd reporting/dash_app
python app.py

and open your browser at http://localhost:8080

Overview

The application contains 3 tabs:

• a first tab showing the distribution of expressions for selected genes, accross of all samples that passed filtering
• a second tab showing the distribution of all gene expressions in selected samples (only for samples that passed filtering)
• a third tab showing scoring / statistics for all genes

The header allows navigating between the tabs. Click on a specific tab to switch to it. The Select data / options button allows selecting the dataset to display, together with other options.

Distribution of gene expressions (Counts / gene)

Users can visualise the distribution of expressions of specific genes.

To select a gene, just start typing its name and click on it when it appears.

Distribution of all gene expressions in specific samples (Counts / sample)

You can visualise the distribution of expressions of all genes in specific samples. Likewise, to select a sample, you may type its name and click on it when it appears.

Gene scoring / statistics (Statistics - all genes)

Scores and statistics for each gene are displayed here.

NB: Users can select a gene directly in the table by ticking the checkbox next to the gene name. This will add the gene to the list of selected genes in the Counts / gene tab.

• merged_data/all_counts.imputed.parquet: parquet file containing all normalised + imputed gene counts
• merged_data/all_counts.parquet: parquet file containing all normalised gene counts
• merged_data/whole_design.csv: table containing designs for all datasets and all samples comprised in the analysis

• 0.downloaded/: raw datasets downloaded from public databases
• 1.id_filtered_renamed/: datasets with filtered and renamed gene IDs
• 2.samples_filtered/: datasets with filtered samples
• 3.: TPM / CPM normalisation
  • 3.tpm_normalised/: TPM normalised datasets
  • 3.cpm_normalised/: CPM normalised datasets
• 4.quantile_normalised/: quantile normalised datasets

• accessions/expression_atlas/: accessions found when querying Expression Atlas
• accessions/geo/: accessions found when querying GEO

• idmapping/global_gene_metadata.csv: table containing the complete set of gene metadata, obtained either via gProfiler or via the custom file provided by the user
• idmapping/global_gene_id_mapping.csv: table containing the complete set of gene id mapping, obtained either via gProfiler or via the custom file
• idmapping/valid_gene_ids.txt: List of gene IDs retained as valid

• gene_transcript_lengths.csv: transcript length relative to each gene ID
• <annotation name>.gff3.gz: GFF3 file

• pipeline_info/
  • Reports generated by Nextflow: execution_report.html, execution_timeline.html, execution_trace.txt and pipeline_dag.dot/pipeline_dag.svg.
  • Reports generated by the pipeline: pipeline_report.html, pipeline_report.txt and software_versions.yml. The pipeline_report* files will only be present if the --email / --email_on_fail parameter’s are used when running the pipeline.
  • Parameters used by the pipeline run: params.json.