clustering_average_similarity {GaMaBioMD}

Perform Hierarchical Clustering for average similarity matrix

Description

This function performs hierarchical clustering based on a similarity matrix and provides additional information such as a colored dendrogram, clustered data, sample percentage within clusters, and total percentage for each cluster.

Usage

clustering_average_similarity(
  similarity_matrix,
  num_clusters_option = 10,
  cut_height_option = 0.2
)

Arguments

similarity_matrix

A matrix containing similarity values between SampleIDs.

num_clusters_option

The desired number of clusters. Default is 10.

cut_height_option

The height at which the dendrogram should be cut to obtain clusters. Default is 0.2.

Value

A list containing the dendrogram, clustered data, sample percentage within clusters, and total percentage for each cluster.

Examples

accession_ranges <- list(
  SRU1 = "AJ240966 to AJ240970",
  STU2 = "AB015240 to AB015245",
  WPU13 = "L11934 to L11939",
  INU20 = c("AF277467 to AF277470", "AF333080 to AF333085")
)

# Use the function to expand accession ranges
sam_acc <- expand_accession_ranges(accession_ranges)
print(sam_acc)

# 2 get_sequence_information
accessions_to_query <- sam_acc$accession
seq_info <- get_sequence_information(accessions_to_query, remove_dot_1 = TRUE)
print(seq_info)
result <- preprocess_for_alignment(sam_acc, seq_info)

# Access the resulting data frames
merged_data <- result$merged_data
main_data <- result$main_data
final_data <- result$final_data

# If you want to sample 10% from each SampleID group:
sampled_data <- data_sampling(final_data, sample_proportion = 0.1)

alignment_results <- alignment_info(final_data, type = "global", verbose = 1)

# Access the resulting data frames
score_matrix <- alignment_results$score_matrix
normalized_score_matrix <- alignment_results$normalized_score_matrix

total_aligned_positions_matrix <- alignment_results$total_aligned_positions_matrix
number_of_matching_positions_matrix <- alignment_results$number_of_matching_positions_matrix

percent_similarity_matrix <- alignment_results$percent_similarity_matrix

alignment_results_list <- alignment_results$alignment_results_list

alignment_info_matrix <- alignment_results$alignment_info_matrix

output_directory <- tempdir()

# Save the list of alignment results to an RDS file
saveRDS(alignment_results_list, file.path(output_directory, "alignment_results_list.rds"))

# Save matrices to files
write.table(score_matrix, file.path(output_directory, "score_matrix.txt"), sep = "\t")
average_percent_similarity <- compute_average_similarity_matrix(percent_similarity_matrix)
print(average_percent_similarity)

output_directory <- tempdir()
width_inch <- 8
height_inch <- 6
dpi <- 300

clustering_result <- clustering_average_similarity(average_percent_similarity)

# Extract the dendrogram and clustered data
dend_colored <- clustering_result$dendrogram
clustered_data <- clustering_result$clustered_data
Cluster_SampleID_Percentage <- clustering_result$Cluster_SampleID_Percentage
Cluster_TotalPercentage <- clustering_result$Cluster_TotalPercentage

tiff_file <- file.path(output_directory, "6. hierarchical_clustering_dendrogram_colored.tiff")

# Save the dendrogram as a TIFF image
tiff(tiff_file, width = width_inch, height = height_inch, units = "in", res = dpi)
plot(dend_colored, main = "Colored Hierarchical Clustering Dendrogram")
dev.off()

# Save the clustered data frame to a CSV file
write.csv(clustered_data, file.path(output_directory, "7. clustered_data.csv"), row.names = FALSE)