23 ARRIVE — animal research
ARRIVE — Animal Research: Reporting of In Vivo Experiments — covers preclinical animal research. The 21-item checklist is at https://arriveguidelines.org/. The templates below cover the statistical analyses most commonly required.
This page contains 50 method-specific templates. Each is a skeleton with placeholders in italics; the prose is written so you can paste it into a manuscript and replace only the bracketed values.
23.0.1 Alignment with BWA and Bowtie
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Alignment with BWA and Bowtie as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Alignment with BWA and Bowtie yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → bwa-bowtie-alignment — For Reviewers
23.0.2 ATAC-Seq Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using ATAC-Seq Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
ATAC-Seq Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → atac-seq-analysis — For Reviewers
23.0.3 Batch Correction with ComBat
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Batch Correction with ComBat as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Batch Correction with ComBat yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → batch-correction-combat — For Reviewers
23.0.4 Bulk RNA-seq Differential Expression with DESeq2
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Bulk RNA-seq Differential Expression with DESeq2 as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Bulk RNA-seq Differential Expression with DESeq2 yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → bulk-rnaseq-differential-expression — For Reviewers
23.0.5 Cell-Type Annotation
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Cell-Type Annotation as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Cell-Type Annotation yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-cell-type-annotation — For Reviewers
23.0.6 ChIP-Seq Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using ChIP-Seq Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
ChIP-Seq Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → chip-seq-analysis — For Reviewers
23.0.7 Copy Number Variation Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Copy Number Variation Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Copy Number Variation Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → cnv-analysis — For Reviewers
23.0.8 Counting Reads with featureCounts
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Counting Reads with featureCounts as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Counting Reads with featureCounts yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → counting-reads-feature-counts — For Reviewers
23.0.9 Differential Expression with edgeR
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Differential Expression with edgeR as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Differential Expression with edgeR yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → edger-differential-expression — For Reviewers
23.0.10 Differential Expression with limma-voom
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Differential Expression with limma-voom as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Differential Expression with limma-voom yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → limma-voom — For Reviewers
23.0.11 DNA Methylation Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using DNA Methylation Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
DNA Methylation Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → methylation-analysis — For Reviewers
23.0.12 Drug-Target Interaction Mining
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Drug-Target Interaction Mining as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Drug-Target Interaction Mining yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → drug-target-interaction — For Reviewers
23.0.13 FASTQ Quality Control
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using FASTQ Quality Control as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
FASTQ Quality Control yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → fastq-quality-control — For Reviewers
23.0.14 Finding Marker Genes
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Finding Marker Genes as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Finding Marker Genes yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-marker-genes — For Reviewers
23.0.15 Gene Annotation with biomaRt
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Gene Annotation with biomaRt as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Gene Annotation with biomaRt yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → biomart-annotation — For Reviewers
23.0.16 Gene Ontology Enrichment
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Gene Ontology Enrichment as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Gene Ontology Enrichment yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → gene-ontology-enrichment — For Reviewers
23.0.17 GSEA Preranked Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using GSEA Preranked Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
GSEA Preranked Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → gsea-preranked — For Reviewers
23.0.18 GSVA Single-Sample Enrichment
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using GSVA Single-Sample Enrichment as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
GSVA Single-Sample Enrichment yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → gsva-single-sample — For Reviewers
23.0.19 Heatmaps for RNA-seq
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Heatmaps for RNA-seq as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Heatmaps for RNA-seq yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → heatmaps-rnaseq — For Reviewers
23.0.20 Integration with Harmony
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Integration with Harmony as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Integration with Harmony yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-integration-harmony — For Reviewers
23.0.21 KEGG Pathway Enrichment
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using KEGG Pathway Enrichment as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
KEGG Pathway Enrichment yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → pathway-enrichment-kegg — For Reviewers
23.0.22 MA Plots
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using MA Plots as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
MA Plots yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → ma-plots — For Reviewers
23.0.23 Metagenomic Profiling
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Metagenomic Profiling as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Metagenomic Profiling yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → metagenomic-profiling — For Reviewers
23.0.24 Microbiome Analysis with DADA2
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Microbiome Analysis with DADA2 as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Microbiome Analysis with DADA2 yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → microbiome-dada2 — For Reviewers
23.0.25 Microbiome Diversity Metrics
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Microbiome Diversity Metrics as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Microbiome Diversity Metrics yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → microbiome-diversity — For Reviewers
23.0.26 Multi-Omics Integration
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Multi-Omics Integration as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Multi-Omics Integration yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → multiomics-integration — For Reviewers
23.0.27 Multiple Sequence Alignment
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Multiple Sequence Alignment as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Multiple Sequence Alignment yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → msa-biostrings — For Reviewers
23.0.28 PCA of RNA-seq Samples
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using PCA of RNA-seq Samples as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
PCA of RNA-seq Samples yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → pca-of-rnaseq — For Reviewers
23.0.29 Phylogenetic Trees with ape
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Phylogenetic Trees with ape as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Phylogenetic Trees with ape yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → phylogenetic-trees-ape — For Reviewers
23.0.30 Population Genetics Basics
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Population Genetics Basics as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Population Genetics Basics yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → population-genetics-basics — For Reviewers
23.0.31 Protein Structure Prediction
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Protein Structure Prediction as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Protein Structure Prediction yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → protein-structure-prediction — For Reviewers
23.0.32 Proteomics with MSstats
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Proteomics with MSstats as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Proteomics with MSstats yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → proteomics-msstats — For Reviewers
23.0.33 Pseudoalignment with Salmon and Kallisto
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Pseudoalignment with Salmon and Kallisto as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Pseudoalignment with Salmon and Kallisto yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → salmon-kallisto-pseudoalignment — For Reviewers
23.0.34 Read Trimming and Adapter Removal
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Read Trimming and Adapter Removal as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Read Trimming and Adapter Removal yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → read-trimming-adapters — For Reviewers
23.0.35 RNA-seq Normalisation
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using RNA-seq Normalisation as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
RNA-seq Normalisation yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → rnaseq-normalization — For Reviewers
23.0.36 scRNA Clustering
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using scRNA Clustering as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
scRNA Clustering yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-clustering — For Reviewers
23.0.37 scRNA Normalisation
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using scRNA Normalisation as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
scRNA Normalisation yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-normalization — For Reviewers
23.0.38 scRNA QC and Filtering
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using scRNA QC and Filtering as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
scRNA QC and Filtering yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-qc-filtering — For Reviewers
23.0.39 Sequence Alignment: Overview
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Sequence Alignment: Overview as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Sequence Alignment: Overview yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → sequence-alignment-overview — For Reviewers
23.0.40 Single-Cell with Seurat
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Single-Cell with Seurat as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Single-Cell with Seurat yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → single-cell-seurat-intro — For Reviewers
23.0.41 Spatial Transcriptomics
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Spatial Transcriptomics as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Spatial Transcriptomics yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → spatial-transcriptomics — For Reviewers
23.0.42 STAR: Spliced Alignment
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using STAR: Spliced Alignment as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
STAR: Spliced Alignment yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → star-spliced-alignment — For Reviewers
23.0.43 Structural Variant Detection
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Structural Variant Detection as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Structural Variant Detection yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → structural-variant-detection — For Reviewers
23.0.44 Surrogate Variable Analysis (SVA)
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Surrogate Variable Analysis (SVA) as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Surrogate Variable Analysis (SVA) yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → surrogate-variable-analysis-sva — For Reviewers
23.0.45 Trajectory Analysis
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Trajectory Analysis as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Trajectory Analysis yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → scrna-trajectory-analysis — For Reviewers
23.0.46 Transcript-to-Gene Summarisation
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Transcript-to-Gene Summarisation as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Transcript-to-Gene Summarisation yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → tx-to-gene-summarization — For Reviewers
23.0.47 Variant Annotation with VEP
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Variant Annotation with VEP as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Variant Annotation with VEP yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → annotation-variant-vep — For Reviewers
23.0.48 Variant Calling with GATK
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Variant Calling with GATK as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Variant Calling with GATK yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → variant-calling-gatk — For Reviewers
23.0.49 VCF Manipulation
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using VCF Manipulation as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
VCF Manipulation yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → vcf-manipulation — For Reviewers
23.0.50 Volcano Plots
Methods paragraph (copy-paste, fill in italics):
[N=] participants/observations were analysed using Volcano Plots as implemented in [package vX.Y] in R [vX.Y]. Assumptions were checked: [list checks performed]. [Effect size with 95% confidence interval reported as the primary outcome]. Two-sided p-values use α = 0.05; [multiple-comparison adjustment if any].
Results paragraph (copy-paste):
Volcano Plots yielded [estimate] (95% CI [lower, upper], p = [value]). [Diagnostic / assumption summary]. [Sensitivity analysis result].
Reviewer checklist for this method. → volcano-plots — For Reviewers
Workflow lab: Goal → Approach → Execution → Check → Report.
23.1 Learning objectives
- Distinguish explanatory from predictive modelling, and choose the correct evaluation metric for each.
- Map STROBE, TRIPOD, STARD, and CONSORT onto the study designs each is meant for.
- Produce a publication-ready regression table with
gtsummarywhose numbers trace directly to the fitted model.
23.3 Background
Breiman’s 1998 essay The Two Cultures and Shmueli’s 2010 essay To Explain or to Predict? draw the same line from opposite sides. In explanatory modelling the goal is inference — what is the estimated effect of X on Y, adjusted for confounders? — and the right evaluation metric is bias-unbiasedness, interval coverage, and interpretability. In predictive modelling the goal is accuracy on unseen data, and the right evaluation metric is out-of-sample loss, calibration, and decision utility. The statistical machinery overlaps but the habits around it do not; a regression that makes an excellent explanation often makes a lacklustre prediction, and vice versa.
Reporting guidelines are the mechanism by which the field enforces discipline around these two cultures. STROBE for observational studies, CONSORT for randomised trials, STARD for diagnostic accuracy, and TRIPOD (now TRIPOD-AI) for prediction models each specify the items a reader needs to evaluate the claim. A checklist filled in as you write is much easier than one filled in at submission.
23.5 1. Goal
Fit a single linear model to the penguins data and present it two ways: once as an explanatory analysis (what drives body mass?) and once as a predictive model (can we predict body mass on held-out birds?).
23.6 2. Approach
fit_expl <- lm(body_mass_g ~ flipper_length_mm + sex + species, data = peng)23.7 3. Execution
peng |>
ggplot(aes(flipper_length_mm, body_mass_g, colour = species)) +
geom_point(alpha = 0.6) +
geom_smooth(method = "lm", se = FALSE, linewidth = 0.8) +
labs(x = "Flipper length (mm)", y = "Body mass (g)")23.8 4. Check
Explanatory evaluation: effect sizes with intervals, residual diagnostics, and adjusted R².
Predictive evaluation: a 5-fold split, hand-coded to keep the example transparent.
folds <- sample(rep(seq_len(k), length.out = nrow(peng)))
rmse_k <- sapply(seq_len(k), function(i) {
tr <- peng[folds != i, ]
te <- peng[folds == i, ]
f <- lm(body_mass_g ~ flipper_length_mm + sex + species, data = tr)
sqrt(mean((predict(f, te) - te$body_mass_g)^2))
})
mean_rmse <- mean(rmse_k)
mean_rmse23.9 5. Report
tbl_regression(fit_expl, intercept = TRUE) |>
modify_caption("**Table 1. Linear-regression estimates for body mass (g).**")In the Palmer penguins dataset (n =
nrow(peng)), body mass was associated with flipper length, sex, and species (adjusted R² =round(summary(fit_expl)$adj.r.squared, 2)). Out-of-sample performance, estimated by 5-fold cross-validation, was RMSE =round(mean_rmse, 0)g.
23.9.1 Reporting-guideline map
| Design | Guideline | URL |
|---|---|---|
| Randomised trial | CONSORT | https://www.consort-statement.org/ |
| Observational study | STROBE | https://www.strobe-statement.org/ |
| Diagnostic-accuracy study | STARD | https://www.equator-network.org/reporting-guidelines/stard/ |
| Prediction-model study | TRIPOD / TRIPOD-AI | https://www.tripod-statement.org/ |
| Systematic review | PRISMA | http://prisma-statement.org/ |
23.10 Common pitfalls
- Choosing a predictor set by R² and reporting a predictive claim, or tuning a predictive model and reporting causal-sounding coefficients.
- Filling in a reporting checklist at submission rather than while drafting.
- Assuming in-sample R² generalises.
23.11 Further reading
- Breiman L. (2001). Statistical Modeling: The Two Cultures.
- Shmueli G. (2010). To Explain or to Predict? Statistical Science.
- Writing a report.