MPhS

With the MPhS package, the user can map their transcriptomic dataset onto the Molecular Phenology Scale (MPhS) proposed by Tornielli GB, Sandri M, Fasoli M, Amato A, Pezzotti M, Zuccolotto P, Zenoni S (2023) A molecular phenology scale of grape berry development. Horticulture Research, Volume 10, Issue 5:uhad048. doi:10.1093/hr/uhad048

Introduction

The molecular phenology scale (MPhS) represents a new tool to precisely align time-series of fruit samples on the basis of molecular changes and to quantify their transcriptomic distance.
The MPhS was built by exploiting molecular-based information from several grape berry transcriptomic datasets.
The proposed statistical pipeline consists of an unsupervised learning procedure yielding an innovative combination of semiparametric, smoothing, and dimensionality reduction tools.
The MPhS is a complementary method for mapping the progression of grape berry development with higher detail compared to classic time- or phenotype-based approaches, and could help coping with challenges such as those raised by climate change.

Installation

You can install the development version of MPhS from GitHub with:

# install.packages("pak")
pak::pak("sndmrc/MPhS")

Example

This is a basic example that shows you how to map the RPKMdata dataset (included in the package) onto the MPhS.
For more information about this dataset, please see the RPKMdata help documentation by using ?RPKMdata.

Load libraries and data.

library(MPhS)
library(tidyr)
library(dplyr)
data("RPKMdata")

The MPhStimepoints command, which maps data to the molecular phenology scale, requires an input dataframe organized with samples as rows and genes as columns, with the following characteristics:
- expression values for each gene must be in separate columns;
- additional columns must be included for experimental conditions and maturation stages;
- each column representing gene expression levels must be named using its gene ID (either V1 or V3 annotation).

Create variables representing the experimental conditions and a variable that defines the maturation stage.

exp_cond <- names(RPKMdata)[-1]
genes <- RPKMdata$gene_id
dts_vars <- data.frame(exp_cond) %>%
   separate(exp_cond, into=c("Cultivar", "Stage", "Replicate"), sep="_")

Transpose the gene expression matrix and add the newly derived variables.

dts <- t(RPKMdata[, -1])
dts <- cbind(dts, dts_vars)
names(dts) <- c(genes, names(dts_vars))

For each stage and each cultivar, calculate the mean value of the 3 replicates (it can takes several minutes).

dts_means <- dts %>%
   group_by(Cultivar, Stage) %>%
   summarize(across(all_of(genes), mean))

Map data onto the transcriptomic scale using the MPhStimepoints command.

MPhS_out <- MPhStimepoints(data=dts_means, strata_var="Cultivar", stage_var="Stage")

The MPhS_out object can be used to visualize the position of the samples on the transcriptomic scale.

p <- plot(MPhS_out)
print(p)