Microbial-Matchmakers

Microbial Matchmakers – Materials and Methods

This project explores how plants regulate microbiome assembly by identifying key transcription factors (TFs) and their gene targets in Arabidopsis thaliana and Lotus japonicus. Below is the full documentation of the bioinformatics pipeline, supporting scripts, and figures.

---

Overview of Methods

1. Gene Selection

Gene expression clusters responsive to synthetic microbial communities (SCs) were extracted from Wippel et al. (2021). Clusters were filtered for transcription factors (TFs) using PlantTFDB annotations. Script used: background_cluster3.py

2. Promoter Sequence Extraction

Arabidopsis promoter sequences were extracted (1000 bp upstream of TSS) using TAIR10. For Lotus, homologous Arabidopsis gene IDs were used due to limited genomic annotation. Scripts used: extract_upstream_promoter_sequences.py, background_cluster3.py

3. Motif Discovery

• FIMO: scanned for known motifs from the JASPAR database
• STREME: discovered novel enriched motifs in promoter sequences Scripts used: Gene_annotation.py, Shuffled_control_At_100_times.py

4. Motif Validation and Enrichment

• GO enrichment using PANTHER
• Motif enrichment statistics using Fisher’s exact test and FDR correction
• 100x shuffled motif controls Scripts used: Motif_distribution_visualization.py, Shuffled_control_At_100_times.py

5. Gene Regulatory Network (GRN) Inference

Used GRNBoost2 (via Arboreto) on transcriptome data from Arabidopsis and Lotus to infer regulatory relationships between TFs and target genes. Scripts used: GRNBoost2_AtSC.py, GRNBoost2_global_GRN_At.py

6. Network Visualization and Robustness Testing

Networks were visualized in Cytoscape. Perturbation analysis simulated TF removals to assess network stability. Scripts used: perturbation_analysis_part_1.py, perturbation_part_2_visualization_plot.py

---

Scripts

Script Name	Description
Gene_annotation.py	Maps FIMO motifs to gene features using GFF3 annotations
GRNBoost2_global_GRN_At.py	Infers a global GRN using expression data across all samples
GRNBoost2_AtSC.py	Infers a GRN using only At-SC samples to highlight context-specific regulation
Motif_distribution_visualization.py	Visualizes positional bias of TF motifs in promoters and computes enrichment
Shuffled_control_At_100_times.py	Performs FIMO scans on 100 randomized promoter sets to calculate empirical p-values
Extract_upstream_promoter_sequences.py	Extracts 1kb upstream sequences from Arabidopsis and Lotus genomes
Background_Arabidopsis_vs_Lotus.py	Creates background gene sets for motif enrichment analysis
perturbation_analysis_part_1.py	Removes individual TFs from GRN and computes network fragmentation
perturbation_part_2_visualization_plot.py	Plots network robustness metrics from perturbation results

---

Figures

Each figure visualizes a key result or method step in the pipeline.

Figure	Description
	Figure S1: Full bioinformatics pipeline overview
	Figure 1: PCA and expression heatmaps from Wippel et al.
	Figure 2: Top TF motifs detected by FIMO
	Figure 3: Statistical enrichment of specific motifs
	Figure 4: Motif significance in Lotus via shuffled control
	Figure 5: Same as above for Arabidopsis
	Figure 6: Differential TF roles in global vs. SC GRNs
	Figure 7: Cytoscape-rendered GRN of Arabidopsis
	Figure 8: Zoom-in on top TFs and their targets
	Figure 9: Effect of TF removals on network fragmentation
	Figure 10: Quantitative summary of perturbation outcomes

---

Data & Resources

• Arabidopsis genome: TAIR10
• Motif database: JASPAR 2022
• Expression data: From Wippel et al. (2021), Nature Microbiology
• Pipeline source repo: MPIPZ_Kathrin_Persistence_RNASeq (GitHub)

---

Citation

If reusing this pipeline, please cite the original data sources and this GitHub repository.