Я пытаюсь найти модули, использующие пакет WGCNA: R, из данных микрочипов для 1500 генов и 34 образцов.

Question

В пакете WGCNA значение R ^ 2 (STF. R.sq) без масштабной топологии должно быть равно 0,8 или выше. Я получаю свой STF. Значение R.sq как 0,4 и ниже. Мой набор данных содержит 34 строки и 1500 столбцов.

     library(WGCNA)
        options(stringsAsFactors = FALSE);

        Alzheimer.1500_series_matrix <- read.csv("C:/Users/Alzheimer.1500_series_matrix.csv", header=FALSE, row.names=1)

        Alzheimer.1500_series_matrix

        # Softthreshold power for noise filtering
        # Choosing the soft-thresholding power co-expression similarity is raised to calculate adjacency
        # By choosing a set of soft-thresholding powers
        powers = c(c(1:22), seq(from = 24, to=34, by=2))

        # pickSoftThreshold performs the analysis of network topology and aids the user in choosing a proper 
          soft-thresholding power

        # To decide scale free topology
        # SFT.R sq value should be above 0.8 for reasonable powers & mean connectivity below 100
        sft = pickSoftThreshold(Alzheimer.1500_series_matrix, powerVector = powers, verbose = 5)

        # Choose the power 6, which is the lowest power for which the scale-free topology index reaches 0.90 as of in tutorial.
        # I am calculating the co expression similarity (adjacencies), using the soft thresholding power 6:

        softPower = 6;
        adjacency = adjacency(Alzheimer.1500_series_matrix, power = softPower);

        # Turn adjacency into topological overlap
        TOM = TOMsimilarity(adjacency); dissTOM = 1-TOM

    # Set the minimum module size relatively high:
    minModuleSize = 30;
# Results of all the modules
Summary

ссылка для получения дополнительной информации о WGCNA