Analysis for Spatiotemporal transcriptomics of water lily (Nymphaea colorata)

Overview

This file contains the complete analysis pipeline for the spatial single‑cell transcriptomics of the tropical water lily (Nymphaea colorata). The project aims to dissect the cellular heterogeneity, developmental trajectories, and molecular mechanisms underlying the formation of complex floral organs (sepals, petals, stamens, and carpels).

Key features

- Single‑cell resolution combined with spatial positional information (barcodes coordinates).- Trajectory inference using Monocle3 to reconstruct organ‑specific differentiation paths.- Focus on MADS‑box tetramer optimisation and cross‑organ expression correlation.- Fully reproducible R‑based workflow with shell scripts for clustering.

Repository structure

stRNA-nym/├── data_for_analysis/ # Intermediate data files│ ├── L7_petal_barcodes_pos.tsv│ ├── L7_sepal_barcodes_pos.tsv│ ├── L7_stamen1-4_barcodes_pos.tsv│ └── ...├── MADS-box_tetramer.R # MADS‑box tetramer optimisation model├── S4_clustering.R # Clustering analysis for Stage4 samples├── cor_nym_vs_pha.R # Cross‑species co‑expression analysis├── cor_ot_vs_it_vs_st.R # Expression correlation across different organs├── meristem_cell_trajectory.R # Monocle trajectory from meristem to organs├── ot2st_trajectory.R # Trajectory from outer tepal to stamen├── steel_clustering.sh # Cell clustering by STEEL├── CITATION.cff└── README.md

Data sources

- Expression matrix – All spatial transcriptomic data are publicly available at:    [http://osf.io/m68cn/overview](http://osf.io/m68cn/overview)- Intermediate analysis files – Pre‑processed barcode position tables and other intermediate files are stored in the [`data_for_analysis/`](data_for_analysis) folder of this repository.

Prerequisites

- R (≥ 4.2) with the following core packages:  ```r  install.packages(c("Seurat", "monocle", "dplyr", "ggplot2", "ggsci", "clustree", "reshape2", "pheatmap"))- STEEL (Spatial Transcriptome based cEll typE cLustering)

STEEL is an unsupervised manifold learning algorithm designed for spatial transcriptome data analysis.

- **Project homepage**: [http://steel-st.sourceforge.io](http://steel-st.sourceforge.io)- **Download**: Get the latest version (source code and precompiled binaries for Linux/macOS) from its [SourceForge page](https://sourceforge.net/projects/steel-st/).- **Installation**:  - Compile from source:    ```bash    g++ src/STEEL.cpp -o steel -O3

Analysis workflow

1. Initial clustering – STEEL and Seurat algorithm applied to spatial barcodes (steel_clustering.sh and S4_clustering.R).2. Trajectory inference – Monocle2 constructs developmental trajectory (see meristem_cell_monocel.R and ot2st_trajectory.R).3. Cross‑organ comparison – Correlation of gene expression between ovary, stamen, and pistil (cor_ot_vs_it_vs_st.R).4. Cross‑species comparison - Correlation of organ co-expression genes between N.colorata and P.aphrodite (cor_nym_vs_pha.R).5. Mechanistic investigation – MADS‑box expression patterns and tetramer optimisation (MADS-box_tetramer.R).