Data files

The folders in this repository contain the data files used in our analysis for the paper "MicroMPN: methods and software for high-throughput screening of microbe suppression in mixed populations" forthcoming in the journal ASM Spectrum.

Authors: Karla Franco Meléndez¹, Layla Schuster², Melinda Chue Donahey², Emily Kairalla², M. Andrew Jansen³, Christopher Reisch^2,4, Adam R. Rivers^1#

Affiliations: ¹ United States Department of Agriculture, Agricultural Research Service, Genomics and Bioinformatics Research Unit, Gainesville, FL 32608, USA; ² Department of Microbiology and Cell Science, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL 32603, USA; ³ United States Department of Agriculture, Agricultural Research Service, Systematic Entomology Laboratory, Electron and Confocal Microscopy Unit, Beltsville, MD 20013, USA; ⁴Present address: Genomatica, San Diego, CA 92121 USA

Data Folders

The data files in this repo contain the raw MPN and CFU counts and their normalized values by grams of soil (MPN/g and CFU/g). All statistical analyses were performed in R v4.3.0 (R Core Team, 2023). Raw and normalized data counts, R outputs, and plots were compiled into Excel spreadsheets.

MPN values were calculated with the Python package "MicroMPN".

1. Folder: plasmid_retention

• This folder contains the CFU/g data corresponding to experiments assessing vector retention in R. solanacearum in soil lacking antibiotic selection. In the Excel files, we referred to data from soil-only microcosms as "simple". Whereas data obtained from soil and synthetic communities were labeled "complex".

  • 1.1. Subfolder: R_analysis

    • This folder contains two files, an R script, and an input file. The R script computes a paired_t_test.

  • 1.2. Subfolder: Excel_spreadsheets

    • This folder contains three Excel files. Raw CFU data sets: "CFU_simple.xlsx" and "CFU_complex.xlsx". The plots published in the paper are found in "data_simple_complex.xlsx".

2. Folder: MicroMPN_soil_only_microcosm

• This folder contains the MPN/g and CFU/g data corresponding to soil-only microcosms experiments.

  • 2.1. Subfolder: micrompn

    • This folder contains the input ("soil_only_input.csv") and output ("OUTFILE_trim_soil_only.csv") files used with "MicroMPN" to calculate MPN values. The output of MicroMPN includes MPN, MPN-adjusted, Jarvis 95% CI, and a rarity index.

  • 2.2. Subfolder: R_analysis

    • This folder contains two files, an R script ("backup_R_corr_regre_BA_ftest_mpn_soil_only.txt") and an input ("soil_only_microcosms.csv") file. The R script computes a Pearson correlation, a linear regression, a Bland-Altman analysis, and Fisher's F-test using MPN/g and CFU/g data. The input file contains both MPN and MPN-adjusted data. Users can run the analysis with MPN-adjusted values by simply subsetting for the correct columns in the input file.

    • Statistical analysis was performed in R v4.3.0 (R Core Team, 2023). Both MPN and MPN-adjusted values were analyzed. However, the scientific paper only discusses and displays results for MPN values. The conclusions of the paper were not affected by MPN-adjusted values.

  • 2.3. Subfolder: Excel_spreadsheet

    • This folder contains a single Excel file ("git_micrompn_trim_norarity_soil_only.xlsx") with the raw data sets, MicroMPN output, CFU/g data, and all of the plots published in the paper.

3. Folder: MicroMPN_soil_and_synthetic_communities

• This folder contains the MPN and CFU count data from soil experiments conducted with synthetic communities. A total of six communities (C1 – C6), each composed of 10 strains, were evaluated for their biocontrol activity against R. solanacearum in soil microcosms.

  • 3.1. Subfolder: micrompn

    • This folder contains the input ("Copy_C1_C2_C3_T48.csv" and "Copy_C4_C5_C6_T48.csv") and output ("OUTFILE_trim_Copy_C1_C2_C3.csv" and "OUTFILE_trim_Copy_C1_C2_C3.csv") files used with "MicroMPN". The output of MicroMPN includes MPN, MPN-adjusted, Jarvis 95% CI, and a rarity index.

  • 3.2. Subfolder: R_analysis

    • This folder contains two files, an R script ("backup_dataMPN_micrompn_trim_norarity_community_R.txt") and an input ("data_micrompn_trim_norarity_community_R.csv") file. The R script computes a Pearson correlation, a linear regression, a Bland-Altman analysis, and Fisher's F-test using MPN/g and CFU/g data. The input file contains both MPN and MPN-adjusted data. Users can run the analysis with MPN-adjusted values by simply subsetting for the correct columns in the input file.

    • Statistical analysis was performed in R v4.3.0 (R Core Team, 2023). Both MPN and MPN-adjusted values were analyzed. However, the scientific paper only discusses and displays results for MPN values. The conclusions of the paper were not affected by MPN-adjusted values.

  • 3.3. Subfolder: Excel_spreadsheet

    • This folder contains a single Excel file ("data_microMPN_TRIM_Improbable.xlsx") with the MicroMPN output, CFU/g data, and all of the plots published in the paper.

  • 3.4. Subfolder: Raw_data

    • This contains two Excel files ("Copy_C1-C3 MPN 48 H MPN fixed.xlsx" and "Copy_C4-C6 MPN 48 H MPN fixed.xlsx") with the raw data sets for MPN and CFU data.

  • 3.5. Subfolder: VCA_analysis

    • This folder contains an R script for running a VCA ("VCA_analysis.4.txt"), input file ("data_micrompn_trim_norarity_community_R.csv"), and several plots from the analysis ("variability_mpn.png", "variability_cfu.png", "vca_plots_normality_outliers.pdf").

4. Folder: multiple_comparisons

• To assess if a community’s suppressive activity was independent of the method (fluorMPN microplate assay vs spotting on agar) being used, we compared the mean of every community to the mean of the control (rfp-Rs5).

  • 4.1. Subfolder: R_analysis

    • This folder contains two files, an R script ("backup_oneway_dataMPN_micrompn_trim_norarity_community_R.txt") and an input ("data_micrompn_trim_norarity_community_R.csv") file. Differences between group means for the fluorMPN microplate assay were compared by a one-way ANOVA, followed by the post-hoc test for multiple comparisons Tukey’s Honest Significant Differences. However, for spotting on agar, since the assumption of normality was violated, we ran the non-parametric test Kruskal-Wallis, followed by Dunn’s test for multiple comparisons. The input file contains both MPN and MPN-adjusted data. Users can run the analysis with MPN-adjusted values by simply subsetting for the correct columns in the input file.

  • 4.2. Subfolder: Excel_spreadsheet

    • This folder contains a single file ("multiple_comparisons.xlsx") with all the plots corresponding to multiple comparison analysis. Both MPN and MPN-adjusted data were analyzed in R and were plotted in Excel.

5. Folder: SpectraMax_output_to_column_format

• This folder contains a Python script ("four_col_updated.py") that converts RFU microplate data (specifically the output of SpectraMax M3) to a column format compatible with "MicroMPN". An example input ("soil_only.txt") and output ("soil_only.csv") files are provided.

  Input Example

  

  Example of SpectraMax M3 RFU table format of a 96-well microplate.

  Output Example

  

  The input data was re-arranged into 4 columns.

6. Folder: microplate_layout

• This folder contains a TOML file ("microplate.toml") specifying a 96-well microplate layout. In our experimental setup, our plate rows (A – H) were assigned a replicate value (1 – 8), while our columns (1 – 12) represented a particular dilution factor (1e00 – 1e-11). For this reason, each plate was divided into a single 12x8 sample block. This TOML file is needed to run "MicroMPN" with any of the micrompn input files provided in this repo.

  TOML Microplate layout example:

#To visualize the TOML file, users will need to first install ["wellmap"](https://wellmap.readthedocs.io/en/latest/index.html). 

wellmap.show("microplate.toml")

To replicate our MPN estimates using the Python package "MicroMPN", the input files (see micrompn folders), and the TOML file provided here, simply run the following command:

! micrompn --wellmap microplate.toml --data "C:\location\of\CSV\file\input_name.csv" --cutoff 6 
--outfile "C:\location\of\CSV\output\file\OUTFILE_name.csv" 
--plate_name "plate_unique" 
--value_name "rfu"
 --well_name "plate_well" 
--trim_positives

• For more information on how to install MicroMPN visit (https://github.com/USDA-ARS-GBRU/micrompn).

• See Jupyter Notebook "Determining_Threshold.ipynb" to learn how to determine a threshold value to supply to MicroMPN.

• An example on how to run MicroMPN from a Jupyter Notebook is available through protocols.io under the name "MicroMPN: Software for Automating Most Probable Number Estimates from Laboratory Microplates" (https://protocols.io/view/micrompn-software-for-automating-most-probable-num-cu5rwy56.html).