Mutanobactin D from the Human Microbiome: Chemistry, Biology and Molecular Dynamics Studies.

This repository contains simulation setups, analysis scripts, and supporting data for the study of Mutanobactin D and its analogs, as described in our associated research paper.

---

📄 Paper

• Link: [To be added upon publication]
• Abstract:
  Mutanobactin D is an interkingdom communicator derived from the human oral microbiome. The lipopeptide prevents yeast-to-hyphae morphogenesis in Candida albicans, notably without fungicidal or fungistatic activity. The mode of action and structure-activity relationship of mutanobactin D are unknown and prompt an interdisciplinary program of study. Stereoselective synthesis of designed mutanobactin D analogs reveals that the C26 configuration is crucial for bioactivity associated with inhibition of pathogenesis, or yeast-to-hyphae transition, in C. albicans. To shed light on this finding, we employ molecular dynamics simulations of mutanobactin D and selected analogs in increasingly complex environments: Monophasic (water or CHCl3), interfacial (water/CHCl3), and explicit lipid membrane (phosphatidylcholine) models. Monophasic MD simulations do not distinguish between bioactive and inactive compounds. In contrast, at a polar/apolar interphase, a dominant, stable conformation emerges for mutanobactin D and bioactive analogs. Explicit lipid membrane simulations reinforce these results and further reveal the formation of a continuous, structured water cushion, which is not found for inactive analogs. Our studies collectively reveal how the stereodefined attachment of the lipid in the C26-C28 motif governs activity against C. albicans and provide a framework for understanding the membrane behavior of mutanobactin D, which may be coupled to its role in the human oral microbiome.

---

📁 Repository Structure

🧪 Simulation Setup

The simulation_setup folder includes simulation configurations for various systems involving the wild-type analog, mutanobactin D (mtb_RR).

Subdirectories:

• 01_monophasic/
• 02_biphasic/
• 03_membrane/

Each contains an example_workflow/ folder with:

• All necessary scripts to replicate the simulations
• 00_master_script.sh to outline the recommended execution order

⚠️ This pipeline is modular, not fully automated, due to manual inspection steps and differing HPC environments.

Requirements:

• gromacs with gmx_mpi
• Tested with:
  • gcc/12.2.0
  • openmpi/4.1.6
  • gromacs/2021.4
• Virtual environment specified in environment.yml

Refer to the paper and supporting information for the exact GROMACS version used in simulations.

---

🧬 Compound Naming

If you encounter analog names in the scripts, they correspond to:

• Prefix mtb → Mutanobactin
• Suffix _* → Analog identifier

Compound/Suffix	Paper ID	Type
RR	1	Active
DOR	7i	Active
KRS	7e	Active
KSS	7d	Active
FRS	C26-b-1	Inactive
DOS	7j	Inactive
KRR	7c	Inactive
KSR	7f	Inactive

---

📊 Analysis

Standard analyses include for example:

• Hydrogen bonding
• Solvent-accessible surface area
• ...

Advanced analysis:

• Membrane-specific metrics (e.g., water cushion)
• Conformational space projection
• Structure refinement to identify cis/trans NOE-constrained states

🧾 Notebooks (analysis/)

• exemplary_analysis_pipeline.ipynb – General trajectory analysis
• membrane_water_cushion_analysis.ipynb – Water cushion quantification
• conformation_projection_plots.ipynb – Projection of analog conformations on i,j vectores

---

📁 Sample Data (data/)

⚠️ Please contact the authors for full trajectories.

Projected coordinates

• .npy arrays of all analysed systems of i, j vectores to reproduce Main Figure 3 via conformation_projection_plots.ipynb
• .csv dataframe of full water cushion analysis of Mutanobactin D WT (1) to run membrane_water_cushion_analysis.ipynb

🧱 Membrane Simulations (Mutanobactin D WT (1) only)

• 1 exemplary membrane simulation trajectory (.pdb, _traj.pdb, .tpr files)
  • **_all_with_water_centered** : mutanobactin, POPC layer and water for water cushion analysis: 4 frames
• Pre-analyzed data:
  • RR_membrane_DICT.pickle: Precalculated results of membrane simulation to run exemplary_analysis_pipeline.ipynb. Structure can be inferred from membrane_extract_traj_features.py
  • coordinates_ij.npy: collected array of i, j vectors of entire trajectory
• EBC clustering files:
  • ebc_nr_clusters_2_temp_100_RR.npy: cluster labels based on entire trajectory via EBC clustering
  • RR_all_pull_cases_pot_ene.npy: Energy data for EBC clustering of entire trajectory

📐 Projected Coordinates

Summary of production simulation descriptors:

Column	Descriptor
1	i-vector
2	j-vector
3	Leucine backbone plane angle
4	Valine angle
5	Lipid tail angle

---