Introduction

This report presents a downstream data analysis and visualization of the matrix obtained from the AgeXtend tool. The matrix contains probability values from various predictive models for a set of chemical compounds. The primary goals of this analysis are to:

  1. Explore the distribution and relationships of anti-aging and “Hallmarks of Aging” probabilities.
  2. Investigate the toxicity and metabolism profiles of the compounds.
  3. Analyze the relationship between predicted probabilities and the Maximum Tolerated Recommended Dose (MTRD).
  4. Identify potential lead compounds with desirable anti-aging properties and acceptable safety profiles.

The dataset includes the following key columns:

Data Loading and Preprocessing

Loading the Data

The first step is to load the data from the provided CSV file.

# Load the data
# Make sure the file 'agextend_probabilities.csv' is in the same directory as this Rmd file,
# or provide the full path to the file.
file_path <- "agextend_probabilities.csv" # Make sure this file is accessible
if (file.exists(file_path)) {
  data_orig <- read.csv(file_path, stringsAsFactors = FALSE, check.names = FALSE)
  data <- data_orig # Work with a copy
  cat("Data loaded successfully.\n")
  cat("Number of rows:", nrow(data), "\n")
  cat("Number of columns:", ncol(data), "\n")
} else {
  warning("The file 'agextend_probabilities.csv' was not found. Using placeholder data for demonstration. Please ensure the file is in the correct path for actual analysis.")
  data <- data.frame(
    SMILES = paste0("C(C\\O)=N#N@", 1:10), # Example SMILES with special chars
    Anti_Aging_Prob = runif(10),
    AIC_Prob1 = runif(10), CS_Prob1 = runif(10), DNS_Prob1 = runif(10),
    EA_Prob1 = runif(10), GI_Prob1 = runif(10), LP_Prob1 = runif(10),
    MD_Prob1 = runif(10), SCE_Prob1 = runif(10), TA_Prob1 = runif(10),
    AMES_Prob1 = runif(10), BBB_Prob1 = runif(10), CYP1A2_Prob1 = runif(10),
    CYP2C19_Prob1 = runif(10), CYP2C9_Prob1 = runif(10), CYP2D6_Prob1 = runif(10),
    CYP3A4_Prob1 = runif(10), DILI_Prob1 = runif(10), Hepato_Prob1 = runif(10),
    hERG_Prob1 = runif(10), HLM_Prob1 = runif(10), MMP_Prob1 = runif(10),
    `P-gp_Inhibitor_Prob1` = runif(10), `P-gp_Substrate_Prob1` = runif(10),
    `MRTD (mg/day)` = rnorm(10, -4, 1),
    `MRTD (uMol)` = rnorm(10, 50, 20),
    check.names = FALSE
  )
}
## Data loaded successfully.
## Number of rows: 120 
## Number of columns: 27
# Display the first few rows using kable for PDF
cat("\nFirst few rows of the dataset:\n")
## 
## First few rows of the dataset:
df_to_print_head <- head(data)
if ("SMILES" %in% colnames(df_to_print_head)) {
  df_to_print_head$SMILES <- sanitize_smiles_for_latex(df_to_print_head$SMILES)
}
kable(df_to_print_head, caption = "First few rows of the dataset.", escape = FALSE)
First few rows of the dataset.
SMILES Anti_Aging_Prob AIC_Prob1 CS_Prob1 DNS_Prob1 EA_Prob1 GI_Prob1 LP_Prob1 MD_Prob1 SCE_Prob1 TA_Prob1 AMES_Prob1 BBB_Prob1 CYP1A2_Prob1 CYP2C19_Prob1 CYP2C9_Prob1 CYP2D6_Prob1 CYP3A4_Prob1 DILI_Prob1 Hepato_Prob1 hERG_Prob1 HLM_Prob1 MMP_Prob1 P-gp_Inhibitor_Prob1 P-gp_Substrate_Prob1 MRTD (mg/day) MRTD (uMol)
0.4153507 0.0870812 0.1679620 0.0022072 0.0233503 0.0000000 0.6721489 0.9549921 0.0470 0.954 0.3541652 0.0012305 0.5704448 0.0503423 0.0108990 0.0540955 0.0460715 0.8058858 0.1062473 0.0709472 0.5895873 0.0169489 0.0240385 0.0870644 -4.422064 37.83864
0.8601621 0.0015100 0.1796195 0.0020561 0.0044521 0.0000000 0.9146299 0.0018636 0.6056 0.758 0.0038247 0.0003003 0.0000862 0.0612354 0.0128114 0.0475504 0.0311646 0.2049868 0.1590862 0.0409310 0.0391542 0.0062550 0.0000000 0.1111346 -3.842360 143.76073
0.8675090 0.0442689 0.1971555 0.0035731 0.0038362 0.0000000 0.5062657 0.4111765 0.6004 0.844 0.8564523 0.0000013 0.0000386 0.0582825 0.0019034 0.0617479 0.0564710 0.5432760 0.1559419 0.0519310 0.1466679 0.0046539 0.0600962 0.0853015 -3.842360 143.76073
0.8675582 0.0104090 0.1560525 0.0065604 0.0023478 0.0092593 0.8536523 0.0187889 0.4318 0.878 0.1791295 0.0000015 0.0008880 0.0639984 0.0087752 0.0486175 0.0328780 0.3559380 0.1664355 0.0435770 0.0929566 0.0034897 0.0000000 0.0580268 -4.422064 37.83864
0.8675303 0.0375370 0.2458189 0.0037018 0.1736523 0.0000000 0.8936273 0.0215177 0.4830 0.900 0.0075567 0.2062904 0.0031630 0.0400446 0.0085117 0.0520556 0.0314254 0.5750942 0.0669128 0.0283625 0.0486844 0.0050056 0.0456731 0.0998663 -4.833956 14.65698
0.5000000 0.0016074 0.0827787 0.0116589 0.0528406 0.0000000 0.9622016 0.0010812 0.3468 0.906 0.1979274 0.0115721 0.0037084 0.0619940 0.0072656 0.0618336 0.0573247 0.6784239 0.1241605 0.0344797 0.1381379 0.0077853 0.0000000 0.2291731 -3.842360 143.76073

Data Cleaning and Preparation

We will perform initial data cleaning, such as checking for missing values and ensuring correct data types. Column names often contain special characters or spaces that can be problematic in R. We will sanitize them.

# Sanitize column names (replace special characters with underscores)
original_colnames <- colnames(data)
colnames(data) <- make.names(colnames(data), unique = TRUE)
sanitized_colnames_df <- data.frame(Original = original_colnames, Sanitized = colnames(data))
cat("Sanitized Column Names:\n")
## Sanitized Column Names:
kable(sanitized_colnames_df, caption = "Mapping of Original to Sanitized Column Names.")
Mapping of Original to Sanitized Column Names.
Original Sanitized
SMILES SMILES
Anti_Aging_Prob Anti_Aging_Prob
AIC_Prob1 AIC_Prob1
CS_Prob1 CS_Prob1
DNS_Prob1 DNS_Prob1
EA_Prob1 EA_Prob1
GI_Prob1 GI_Prob1
LP_Prob1 LP_Prob1
MD_Prob1 MD_Prob1
SCE_Prob1 SCE_Prob1
TA_Prob1 TA_Prob1
AMES_Prob1 AMES_Prob1
BBB_Prob1 BBB_Prob1
CYP1A2_Prob1 CYP1A2_Prob1
CYP2C19_Prob1 CYP2C19_Prob1
CYP2C9_Prob1 CYP2C9_Prob1
CYP2D6_Prob1 CYP2D6_Prob1
CYP3A4_Prob1 CYP3A4_Prob1
DILI_Prob1 DILI_Prob1
Hepato_Prob1 Hepato_Prob1
hERG_Prob1 hERG_Prob1
HLM_Prob1 HLM_Prob1
MMP_Prob1 MMP_Prob1
P-gp_Inhibitor_Prob1 P.gp_Inhibitor_Prob1
P-gp_Substrate_Prob1 P.gp_Substrate_Prob1
MRTD (mg/day) MRTD..mg.day.
MRTD (uMol) MRTD..uMol. 
# Identify column groups based on sanitized names
# These indices/names must match your actual dataset structure after sanitization
anti_aging_col <- colnames(data)[2]
hallmarks_cols <- colnames(data)[3:11]
toxicity_metabolism_cols <- colnames(data)[12:25]
mrtd_mg_col <- colnames(data)[26]
mrtd_umol_col <- colnames(data)[27]

# Ensure the identified columns exist
# Adding checks to prevent errors if columns are not as expected
expected_col_indices <- c(2, 3:11, 12:25, 26, 27)
if (max(expected_col_indices) > ncol(data)) {
  stop("The dataset does not have the expected number of columns. Please check the input file and column definitions.")
}

stopifnot(anti_aging_col %in% colnames(data))
stopifnot(all(hallmarks_cols %in% colnames(data)))
stopifnot(all(toxicity_metabolism_cols %in% colnames(data)))
stopifnot(mrtd_mg_col %in% colnames(data))
stopifnot(mrtd_umol_col %in% colnames(data))

probability_cols <- c(anti_aging_col, hallmarks_cols, toxicity_metabolism_cols)

# Convert probability columns to numeric
for (col in probability_cols) {
  if (!is.numeric(data[[col]])) {
    data[[col]] <- as.numeric(as.character(data[[col]]))
  }
}

# Convert MRTD columns to numeric
if (!is.numeric(data[[mrtd_mg_col]])) {
  data[[mrtd_mg_col]] <- as.numeric(as.character(data[[mrtd_mg_col]]))
}
if (!is.numeric(data[[mrtd_umol_col]])) {
  data[[mrtd_umol_col]] <- as.numeric(as.character(data[[mrtd_umol_col]]))
}

# Check for missing values
missing_values <- sapply(data, function(x) sum(is.na(x)))
missing_summary <- data.frame(Column = names(missing_values), MissingCount = missing_values)
missing_summary <- missing_summary[missing_summary$MissingCount > 0, ]

if (nrow(missing_summary) > 0) {
  cat("\nColumns with missing values:\n")
  kable(missing_summary, caption = "Missing Value Summary.")
  warning("Missing values detected. Consider imputation or removal.")
} else {
  cat("\nNo missing values found in the dataset.\n")
}
## 
## No missing values found in the dataset.
cat("\nData types of columns:\n")
## 
## Data types of columns:
kable(data.frame(Column = colnames(data), DataType = sapply(data, class)), caption = "Column Data Types.")
Column Data Types.
Column DataType
SMILES SMILES character
Anti_Aging_Prob Anti_Aging_Prob numeric
AIC_Prob1 AIC_Prob1 numeric
CS_Prob1 CS_Prob1 numeric
DNS_Prob1 DNS_Prob1 numeric
EA_Prob1 EA_Prob1 numeric
GI_Prob1 GI_Prob1 numeric
LP_Prob1 LP_Prob1 numeric
MD_Prob1 MD_Prob1 numeric
SCE_Prob1 SCE_Prob1 numeric
TA_Prob1 TA_Prob1 numeric
AMES_Prob1 AMES_Prob1 numeric
BBB_Prob1 BBB_Prob1 numeric
CYP1A2_Prob1 CYP1A2_Prob1 numeric
CYP2C19_Prob1 CYP2C19_Prob1 numeric
CYP2C9_Prob1 CYP2C9_Prob1 numeric
CYP2D6_Prob1 CYP2D6_Prob1 numeric
CYP3A4_Prob1 CYP3A4_Prob1 numeric
DILI_Prob1 DILI_Prob1 numeric
Hepato_Prob1 Hepato_Prob1 numeric
hERG_Prob1 hERG_Prob1 numeric
HLM_Prob1 HLM_Prob1 numeric
MMP_Prob1 MMP_Prob1 numeric
P.gp_Inhibitor_Prob1 P.gp_Inhibitor_Prob1 numeric
P.gp_Substrate_Prob1 P.gp_Substrate_Prob1 numeric
MRTD..mg.day. MRTD..mg.day. numeric
MRTD..uMol. MRTD..uMol. numeric

The data has been loaded and initial preprocessing steps, including column name sanitization and data type conversion, have been performed. We’ve also checked for missing values.

Exploratory Data Analysis (EDA)

Summary Statistics

Let’s look at the summary statistics for the key probability columns and MTRD values.

# Summary for Anti-Aging Probability
cat("\nSummary for Anti-Aging Probability (", anti_aging_col, "):\n")
## 
## Summary for Anti-Aging Probability ( Anti_Aging_Prob ):
summary_anti_aging <- summary(data[[anti_aging_col]])
print(summary_anti_aging) 
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##  0.2072  0.8299  0.8691  0.8238  0.9131  0.9607
# Summary for Hallmarks of Aging Probabilities
cat("\nSummary for Hallmarks of Aging Probabilities:\n")
## 
## Summary for Hallmarks of Aging Probabilities:
summary_hallmarks_df <- as.data.frame(t(as.matrix(summary(data[, hallmarks_cols]))))
kable(summary_hallmarks_df, caption = "Summary Statistics for Hallmarks of Aging Probabilities.")
Summary Statistics for Hallmarks of Aging Probabilities.
Var1 Var2 Freq
AIC_Prob1 Min. :0.0002661
CS_Prob1 Min. :0.08278
DNS_Prob1 Min. :0.0002683
EA_Prob1 Min. :0.0009009
GI_Prob1 Min. :0.000000
LP_Prob1 Min. :0.1344
MD_Prob1 Min. :0.0002777
SCE_Prob1 Min. :0.0470
TA_Prob1 Min. :0.2640
AIC_Prob1 1st Qu.:0.0172949
CS_Prob1 1st Qu.:0.17746
DNS_Prob1 1st Qu.:0.0028486
EA_Prob1 1st Qu.:0.0035450
GI_Prob1 1st Qu.:0.000000
LP_Prob1 1st Qu.:0.6324
MD_Prob1 1st Qu.:0.0032534
SCE_Prob1 1st Qu.:0.3944
TA_Prob1 1st Qu.:0.6915
AIC_Prob1 Median :0.0912741
CS_Prob1 Median :0.23274
DNS_Prob1 Median :0.0067976
EA_Prob1 Median :0.0044528
GI_Prob1 Median :0.000000
LP_Prob1 Median :0.7949
MD_Prob1 Median :0.0191052
SCE_Prob1 Median :0.4936
TA_Prob1 Median :0.7820
AIC_Prob1 Mean :0.2797469
CS_Prob1 Mean :0.29327
DNS_Prob1 Mean :0.1225560
EA_Prob1 Mean :0.0721175
GI_Prob1 Mean :0.002186
LP_Prob1 Mean :0.7371
MD_Prob1 Mean :0.1220043
SCE_Prob1 Mean :0.4907
TA_Prob1 Mean :0.7649
AIC_Prob1 3rd Qu.:0.5041392
CS_Prob1 3rd Qu.:0.38338
DNS_Prob1 3rd Qu.:0.0222361
EA_Prob1 3rd Qu.:0.0103979
GI_Prob1 3rd Qu.:0.000000
LP_Prob1 3rd Qu.:0.8857
MD_Prob1 3rd Qu.:0.1225927
SCE_Prob1 3rd Qu.:0.6025
TA_Prob1 3rd Qu.:0.8885
AIC_Prob1 Max. :0.9962562
CS_Prob1 Max. :0.76162
DNS_Prob1 Max. :0.9971994
EA_Prob1 Max. :1.0000000
GI_Prob1 Max. :0.064815
LP_Prob1 Max. :0.9855
MD_Prob1 Max. :0.9722046
SCE_Prob1 Max. :0.9358
TA_Prob1 Max. :1.0000 
# Summary for Toxicity and Metabolism Probabilities
cat("\nSummary for Toxicity and Metabolism Probabilities:\n")
## 
## Summary for Toxicity and Metabolism Probabilities:
summary_tox_met_df <- as.data.frame(t(as.matrix(summary(data[, toxicity_metabolism_cols]))))
kable(summary_tox_met_df, caption = "Summary Statistics for Toxicity and Metabolism Probabilities.")
Summary Statistics for Toxicity and Metabolism Probabilities.
Var1 Var2 Freq
AMES_Prob1 Min. :6.050e-06
BBB_Prob1 Min. :4.000e-08
CYP1A2_Prob1 Min. :9.000e-08
CYP2C19_Prob1 Min. :0.01416
CYP2C9_Prob1 Min. :0.001778
CYP2D6_Prob1 Min. :0.04077
CYP3A4_Prob1 Min. :0.02139
DILI_Prob1 Min. :0.0879
Hepato_Prob1 Min. :0.03889
hERG_Prob1 Min. :0.02836
HLM_Prob1 Min. :0.003325
MMP_Prob1 Min. :0.002182
P.gp_Inhibitor_Prob1 Min. :0.00000
P.gp_Substrate_Prob1 Min. :0.05803
AMES_Prob1 1st Qu.:2.942e-03
BBB_Prob1 1st Qu.:1.790e-06
CYP1A2_Prob1 1st Qu.:4.595e-05
CYP2C19_Prob1 1st Qu.:0.05406
CYP2C9_Prob1 1st Qu.:0.007856
CYP2D6_Prob1 1st Qu.:0.05197
CYP3A4_Prob1 1st Qu.:0.03908
DILI_Prob1 1st Qu.:0.2882
Hepato_Prob1 1st Qu.:0.12593
hERG_Prob1 1st Qu.:0.04394
HLM_Prob1 1st Qu.:0.048662
MMP_Prob1 1st Qu.:0.005662
P.gp_Inhibitor_Prob1 1st Qu.:0.00000
P.gp_Substrate_Prob1 1st Qu.:0.09943
AMES_Prob1 Median :3.526e-02
BBB_Prob1 Median :2.625e-05
CYP1A2_Prob1 Median :4.387e-04
CYP2C19_Prob1 Median :0.06117
CYP2C9_Prob1 Median :0.012958
CYP2D6_Prob1 Median :0.05788
CYP3A4_Prob1 Median :0.04641
DILI_Prob1 Median :0.4574
Hepato_Prob1 Median :0.18109
hERG_Prob1 Median :0.05959
HLM_Prob1 Median :0.111367
MMP_Prob1 Median :0.016341
P.gp_Inhibitor_Prob1 Median :0.01442
P.gp_Substrate_Prob1 Median :0.14518
AMES_Prob1 Mean :2.344e-01
BBB_Prob1 Mean :1.656e-01
CYP1A2_Prob1 Mean :8.967e-02
CYP2C19_Prob1 Mean :0.11475
CYP2C9_Prob1 Mean :0.083196
CYP2D6_Prob1 Mean :0.06372
CYP3A4_Prob1 Mean :0.07823
DILI_Prob1 Mean :0.4595
Hepato_Prob1 Mean :0.19746
hERG_Prob1 Mean :0.07897
HLM_Prob1 Mean :0.198558
MMP_Prob1 Mean :0.199316
P.gp_Inhibitor_Prob1 Mean :0.16521
P.gp_Substrate_Prob1 Mean :0.26951
AMES_Prob1 3rd Qu.:3.566e-01
BBB_Prob1 3rd Qu.:8.426e-03
CYP1A2_Prob1 3rd Qu.:7.214e-03
CYP2C19_Prob1 3rd Qu.:0.07844
CYP2C9_Prob1 3rd Qu.:0.096712
CYP2D6_Prob1 3rd Qu.:0.06174
CYP3A4_Prob1 3rd Qu.:0.05838
DILI_Prob1 3rd Qu.:0.6586
Hepato_Prob1 3rd Qu.:0.25466
hERG_Prob1 3rd Qu.:0.08283
HLM_Prob1 3rd Qu.:0.272523
MMP_Prob1 3rd Qu.:0.310225
P.gp_Inhibitor_Prob1 3rd Qu.:0.16587
P.gp_Substrate_Prob1 3rd Qu.:0.25494
AMES_Prob1 Max. :9.987e-01
BBB_Prob1 Max. :9.997e-01
CYP1A2_Prob1 Max. :9.446e-01
CYP2C19_Prob1 Max. :0.91341
CYP2C9_Prob1 Max. :0.779949
CYP2D6_Prob1 Max. :0.65029
CYP3A4_Prob1 Max. :0.91251
DILI_Prob1 Max. :0.8059
Hepato_Prob1 Max. :0.91012
hERG_Prob1 Max. :0.42637
HLM_Prob1 Max. :0.873926
MMP_Prob1 Max. :0.973326
P.gp_Inhibitor_Prob1 Max. :1.00000
P.gp_Substrate_Prob1 Max. :0.93609 
# Summary for MRTD values
cat("\nSummary for MRTD (", mrtd_mg_col, "):\n")
## 
## Summary for MRTD ( MRTD..mg.day. ):
summary_mrtd_mg <- summary(data[[mrtd_mg_col]])
print(summary_mrtd_mg)
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
##  -6.334  -5.148  -4.422  -4.589  -3.842  -3.842
cat("\nSummary for MRTD (", mrtd_umol_col, "):\n")
## 
## Summary for MRTD ( MRTD..uMol. ):
summary_mrtd_umol <- summary(data[[mrtd_umol_col]])
print(summary_mrtd_umol)
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
##   0.4631   7.1163  37.8386  57.7677 143.7607 143.7607

The summary statistics provide an overview of the central tendency, spread, and range of the predicted probabilities and MTRD values.

Distribution of Probabilities

Visualizing the distribution of probabilities can reveal patterns and outliers.

Anti-Aging Probability

ggplot(data, aes_string(x = anti_aging_col)) +
  geom_histogram(binwidth = 0.05, fill = "steelblue", color = "black", alpha = 0.7) +
  labs(title = "Distribution of Anti-Aging Probability",
       x = "Anti-Aging Probability",
       y = "Frequency") +
  custom_theme()
Distribution of Anti-Aging Probability.

Distribution of Anti-Aging Probability. 

ggplot(data, aes_string(y = anti_aging_col)) + 
  geom_boxplot(fill = "skyblue", color = "black", alpha = 0.7, width=0.3) +
  labs(title = "Boxplot of Anti-Aging Probability",
       y = "Anti-Aging Probability", x="") + 
  custom_theme() +
  theme(axis.text.x = element_blank(), axis.ticks.x = element_blank())
Distribution of Anti-Aging Probability.

Distribution of Anti-Aging Probability.

The histogram and boxplot show the distribution of the Anti-Aging probabilities.

Hallmarks of Aging Probabilities

data_hallmarks_long <- data %>%
  select(all_of(hallmarks_cols)) %>%
  pivot_longer(everything(), names_to = "Hallmark", values_to = "Probability")

ggplot(data_hallmarks_long, aes(x = Probability)) +
  geom_histogram(binwidth = 0.05, fill = "coral", color = "black", alpha = 0.7) +
  facet_wrap(~ Hallmark, scales = "free_y", ncol = 3) +
  labs(title = "Distributions of Hallmarks of Aging Probabilities",
       x = "Probability",
       y = "Frequency") +
  custom_theme() +
  theme(strip.text = element_text(size = 7))
Distributions of Hallmarks of Aging Probabilities.

Distributions of Hallmarks of Aging Probabilities. 

ggplot(data_hallmarks_long, aes(x = Hallmark, y = Probability, fill = Hallmark)) +
  geom_boxplot(alpha = 0.7, show.legend = FALSE) +
  labs(title = "Boxplots of Hallmarks of Aging Probabilities",
       x = "Hallmark of Aging",
       y = "Probability") +
  custom_theme() +
  theme(axis.text.x = element_text(angle = 60, hjust = 1, size=7))
Distributions of Hallmarks of Aging Probabilities.

Distributions of Hallmarks of Aging Probabilities.

These plots display the distributions for each of the 9 “Hallmarks of Aging” probabilities.

Toxicity and Metabolism Probabilities

data_tox_met_long <- data %>%
  select(all_of(toxicity_metabolism_cols)) %>%
  pivot_longer(everything(), names_to = "ToxMet_Property", values_to = "Probability")

ggplot(data_tox_met_long, aes(x = Probability)) +
  geom_histogram(binwidth = 0.05, fill = "lightgreen", color = "black", alpha = 0.7) +
  facet_wrap(~ ToxMet_Property, scales = "free_y", ncol = 3) +
  labs(title = "Distributions of Toxicity and Metabolism Probabilities",
       x = "Probability",
       y = "Frequency") +
  custom_theme() +
  theme(strip.text = element_text(size = 6))
Distributions of Toxicity and Metabolism Probabilities.

Distributions of Toxicity and Metabolism Probabilities. 

ggplot(data_tox_met_long, aes(x = ToxMet_Property, y = Probability, fill = ToxMet_Property)) +
  geom_boxplot(alpha = 0.7, show.legend = FALSE) +
  labs(title = "Boxplots of Toxicity and Metabolism Probabilities",
       x = "Toxicity/Metabolism Property",
       y = "Probability") +
  custom_theme() +
  theme(axis.text.x = element_text(angle = 75, hjust = 1, size=6))
Distributions of Toxicity and Metabolism Probabilities.

Distributions of Toxicity and Metabolism Probabilities.

These visualizations show the distributions for various toxicity and metabolism-related probabilities.

Correlation Analysis

Correlation Matrix of All Probabilities

probability_data <- data[, probability_cols]
cor_matrix <- cor(probability_data, use = "pairwise.complete.obs")

corrplot(cor_matrix,
         method = "color",
         type = "upper",
         order = "hclust",
         tl.col = "black", tl.srt = 45, tl.cex = 0.5, 
         addCoef.col = "black", number.cex = 0.4, 
         col = brewer.pal(n = 8, name = "RdYlBu"),
         diag = FALSE,
         title = "Correlation Plot of All Probabilities", mar=c(0,0,1,0))
Correlation Matrix of All Predicted Probabilities.

Correlation Matrix of All Predicted Probabilities.

The correlation matrix visualizes pairwise Pearson correlations.

Correlation between Anti-Aging and Hallmarks

anti_aging_hallmarks_data <- data[, c(anti_aging_col, hallmarks_cols)]
cor_matrix_ah <- cor(anti_aging_hallmarks_data, use = "pairwise.complete.obs")

corrplot(cor_matrix_ah, method = "number", type = "upper", order = "original",
         tl.col = "black", tl.srt = 45, tl.cex = 0.7, 
         number.cex = 0.7, 
         title = "Correlations: Anti-Aging vs. Hallmarks", mar=c(0,0,1,0))
Correlations: Anti-Aging vs. Hallmarks.

Correlations: Anti-Aging vs. Hallmarks.

This plot focuses on correlations between Anti-Aging probability and individual Hallmarks of Aging.

Compound Profiling and Clustering

Heatmap of Probabilities

num_compounds_heatmap <- min(nrow(data), 30) 
# Ensure probability_data is defined if not run globally
if (!exists("probability_data")) {
    probability_data <- data[, probability_cols]
}
heatmap_data_subset <- probability_data[1:num_compounds_heatmap, ]

# Use short identifiers for rownames if SMILES are too long for PDF
# Use original SMILES from 'data_orig' or 'data' before sanitization for actual SMILES if needed for other purposes
# but for rownames in heatmap, use sanitized or placeholder
short_ids <- if ("SMILES" %in% colnames(data) && num_compounds_heatmap > 0) {
                 data$SMILES[1:num_compounds_heatmap] # Use the sanitized SMILES from the 'data' df
             } else if (num_compounds_heatmap > 0) {
                 rownames(heatmap_data_subset)
             } else {
                 character(0)
             }
# Sanitize these short_ids if they are SMILES, for display in heatmap (pheatmap might handle some)
# However, pheatmap usually shows rownames as is.
# For safety, if these are actual SMILES, they might need sanitization for LaTeX context if pheatmap generates LaTeX code.
# But pheatmap generates a plot, not LaTeX table code. So, raw SMILES as rownames should be fine here.
if (length(short_ids) == nrow(heatmap_data_subset)) { # Check if short_ids generation was successful
    rownames(heatmap_data_subset) <- short_ids
}


pheatmap(heatmap_data_subset,
         scale = "none",
         clustering_distance_rows = "euclidean",
         clustering_distance_cols = "euclidean",
         clustering_method = "ward.D2",
         cutree_rows = if(num_compounds_heatmap > 1) min(3, num_compounds_heatmap-1) else 1,
         cutree_cols = 2,
         fontsize_row = 5, 
         fontsize_col = 6, 
         angle_col = "45",
         main = paste("Heatmap for first", num_compounds_heatmap, "Compounds"),
         color = colorRampPalette(rev(brewer.pal(n = 7, name = "RdYlBu")))(100))
Heatmap of Probabilities.

Heatmap of Probabilities.

The heatmap clusters compounds and probability models. Note: Displays a subset of compounds.

K-Means Clustering

clustering_data_kmeans <- data[, c(anti_aging_col, hallmarks_cols)]
clustering_data_kmeans <- na.omit(clustering_data_kmeans) 
if(nrow(clustering_data_kmeans) < 2) { # Changed from clustering_data to clustering_data_kmeans
  cat("Not enough data points for clustering after NA removal.\n")
} else {
  clustering_data_scaled <- scale(clustering_data_kmeans) # Changed

  cat("Optimal k determination plots omitted for brevity in PDF. Run interactively if needed.\n")
  
  k_optimal <- 3 
  if(nrow(clustering_data_scaled) < k_optimal) k_optimal <- nrow(clustering_data_scaled)
  if(k_optimal == 0 && nrow(clustering_data_scaled) > 0) k_optimal <- 1 # Handle edge case for very few rows
  
  if(k_optimal > 0) { # Proceed only if k_optimal is valid
    set.seed(123)
    kmeans_result <- kmeans(clustering_data_scaled, centers = k_optimal, nstart = 25)
    
    data_for_pca <- data[rownames(clustering_data_kmeans), ] 
    data_for_pca$Cluster_Hallmarks <- as.factor(kmeans_result$cluster)
    
    pca_result <- prcomp(clustering_data_scaled, center = TRUE, scale. = TRUE)
    print(fviz_pca_ind(pca_result, # Explicitly print ggplot object
                 geom.ind = "point",
                 col.ind = data_for_pca$Cluster_Hallmarks,
                 palette = "jco",
                 addEllipses = TRUE, ellipse.type = "confidence",
                 legend.title = "Clusters",
                 title = "PCA of Compounds (Anti-Aging & Hallmarks)") + custom_theme())
    
    cat("\nSummary of compounds per cluster (based on Hallmarks):\n")
    kable(table(data_for_pca$Cluster_Hallmarks), col.names = c("Cluster", "Count"), caption = "Compound Counts per Cluster.")
  } else {
    cat("Cannot perform K-Means clustering with k_optimal <= 0.\n")
  }
}
## Optimal k determination plots omitted for brevity in PDF. Run interactively if needed.
K-Means Clustering of Compounds based on Hallmarks.

K-Means Clustering of Compounds based on Hallmarks. 

## 
## Summary of compounds per cluster (based on Hallmarks):
Compound Counts per Cluster.
Cluster Count
1 23
2 27
3 70

K-Means clustering partitions compounds. PCA visualizes these clusters.

Analysis of Toxicity and Metabolism

Highlighting Compounds with Potential Toxicity Issues

toxicity_threshold <- 0.7
specific_toxicity_cols_pattern <- "AMES|DILI|Hepato|hERG"
# Use original toxicity_metabolism_cols for grep, as it's based on original column structure assumptions
specific_toxicity_cols <- grep(specific_toxicity_cols_pattern, colnames(data), value = TRUE, ignore.case = TRUE)
# Filter to ensure these are indeed from the intended block of columns
specific_toxicity_cols <- intersect(specific_toxicity_cols, toxicity_metabolism_cols)


if (length(specific_toxicity_cols) > 0 && all(specific_toxicity_cols %in% colnames(data))) {
  high_toxicity_matrix <- data[, specific_toxicity_cols, drop=FALSE] > toxicity_threshold
  data$Num_High_Toxicity_Flags <- rowSums(high_toxicity_matrix, na.rm = TRUE)
  
  cat("\nDistribution of Number of High Toxicity Flags (Threshold > ", toxicity_threshold, "):\n")
  print(table(data$Num_High_Toxicity_Flags))
  
  # Select SMILES from original data (data_orig) if available, or current data
  smiles_col_name <- names(data_orig)[1] # Assuming first col is SMILES in original
  if (!("SMILES" %in% colnames(data))) data$SMILES <- data_orig[[smiles_col_name]]


  compounds_with_high_toxicity <- data[data$Num_High_Toxicity_Flags > 0, c("SMILES", anti_aging_col, specific_toxicity_cols, "Num_High_Toxicity_Flags")]
  
  cat("\nCompounds with at least one high toxicity probability (Threshold > ", toxicity_threshold, "):\n")
  if (nrow(compounds_with_high_toxicity) > 0) {
    if ("SMILES" %in% colnames(compounds_with_high_toxicity)) {
      compounds_with_high_toxicity$SMILES <- sanitize_smiles_for_latex(compounds_with_high_toxicity$SMILES)
    }
    kable(compounds_with_high_toxicity, caption = paste("Compounds with High Toxicity Flags (Prob >", toxicity_threshold,")."), escape = FALSE)
  } else {
    cat("No compounds found with high toxicity flags above the threshold.\n")
  }
} else {
  cat("No specific toxicity columns found matching AMES, DILI, Hepato, hERG patterns or they are not in dataframe. Skipping this section.\n")
}
## 
## Distribution of Number of High Toxicity Flags (Threshold >  0.7 ):
## 
##  0  1  2 
## 82 32  6 
## 
## Compounds with at least one high toxicity probability (Threshold >  0.7 ):
Compounds with High Toxicity Flags (Prob > 0.7 ).
SMILES Anti_Aging_Prob AMES_Prob1 DILI_Prob1 Hepato_Prob1 hERG_Prob1 Num_High_Toxicity_Flags
1 0.4153507 0.3541652 0.8058858 0.1062473 0.0709472 1
3 0.8675090 0.8564523 0.5432760 0.1559419 0.0519310 1
7 0.8134314 0.9755830 0.6875242 0.0980618 0.0316115 1
13 0.8845954 0.0561883 0.7032266 0.0877149 0.0521452 1
14 0.9026542 0.9954975 0.4241290 0.0953081 0.0827177 1
16 0.8674786 0.1039045 0.7423505 0.9101161 0.2024326 2
18 0.9066839 0.8257204 0.2085714 0.2077957 0.0696443 1
29 0.8690741 0.8275343 0.6846678 0.1842713 0.0573852 1
31 0.8772374 0.1013908 0.7551078 0.0991374 0.0406885 1
33 0.3721187 0.9938579 0.2888746 0.1843909 0.0331029 1
36 0.7298995 0.0501221 0.7141211 0.1311200 0.0607621 1
37 0.8675245 0.0273033 0.8018489 0.0511540 0.0484780 1
39 0.8422391 0.9978563 0.7096740 0.1136693 0.0393497 2
40 0.9339135 0.9514049 0.7388092 0.0388904 0.0614048 2
43 0.9068187 0.0030837 0.7131538 0.1783630 0.0596383 1
47 0.9213936 0.9496986 0.2405188 0.1720684 0.1071536 1
53 0.9522983 0.7597901 0.6579833 0.0877450 0.0521926 1
54 0.8674711 0.8817641 0.1550104 0.2075963 0.0556442 1
56 0.7576527 0.8118979 0.1186264 0.2327908 0.0904809 1
57 0.8675203 0.8689586 0.1803063 0.2515460 0.0579422 1
61 0.9028072 0.7229171 0.1823889 0.2739984 0.0467560 1
68 0.9147427 0.9972081 0.1696082 0.2545630 0.1695946 1
69 0.9147427 0.9972081 0.1696082 0.2545630 0.1695946 1
77 0.6541684 0.7139102 0.6066611 0.2645012 0.0391153 1
79 0.8674408 0.8867352 0.3191527 0.2622778 0.0508184 1
81 0.8309126 0.0123644 0.7163282 0.1081725 0.0308414 1
84 0.7383333 0.0075317 0.7585896 0.2179448 0.1036391 1
86 0.9113007 0.8626630 0.7229768 0.1576709 0.0401219 2
88 0.9358224 0.9195161 0.6822173 0.0468319 0.0584787 1
89 0.8745806 0.2100843 0.7459675 0.7528580 0.1423502 2
90 0.8675542 0.0055737 0.7233558 0.0877022 0.0452627 1
93 0.8819882 0.0104895 0.8025137 0.2549616 0.3522341 1
94 0.8885711 0.7305439 0.7487818 0.0703219 0.0451110 2
99 0.9362787 0.9986765 0.6636929 0.3792285 0.0351686 1
102 0.8087053 0.0028934 0.7678359 0.1265243 0.1031895 1
103 0.8675542 0.0055737 0.7233558 0.0877022 0.0452627 1
117 0.9141520 0.9216489 0.2223560 0.2563339 0.0521820 1
119 0.8708766 0.0697755 0.7433143 0.1443709 0.0373102 1

This section identifies compounds with high predicted toxicity.

Metabolism Profile Overview (e.g., CYP Interaction)

cyp_cols_pattern <- "CYP"
cyp_cols <- grep(cyp_cols_pattern, colnames(data), value = TRUE, ignore.case = TRUE)
cyp_cols <- intersect(cyp_cols, toxicity_metabolism_cols)


if (length(cyp_cols) > 0 && all(cyp_cols %in% colnames(data))) {
  # Ensure SMILES column is present for pivot_longer if it's used as id_cols implicitly
  # It's better to explicitly select columns for pivot_longer
  data_cyp_long <- data %>%
    select(all_of(cyp_cols)) %>% # Select only CYP columns for pivoting
    pivot_longer(cols = everything(), names_to = "CYP_Enzyme", values_to = "Probability")
  
  print(ggplot(data_cyp_long, aes(x = CYP_Enzyme, y = Probability, fill = CYP_Enzyme)) + # Explicitly print
    geom_boxplot(show.legend = FALSE, alpha = 0.7) +
    labs(title = "CYP Interaction Probabilities",
         x = "CYP Model",
         y = "Probability") +
    custom_theme() +
    theme(axis.text.x = element_text(angle = 60, hjust = 1, size = 7)))
  
  cyp_threshold <- 0.7
  high_cyp_matrix <- data[, cyp_cols, drop=FALSE] > cyp_threshold
  data$Num_High_CYP_Flags <- rowSums(high_cyp_matrix, na.rm = TRUE)
  
  # Ensure SMILES column exists in data for selection
  if (!("SMILES" %in% colnames(data)) && ("SMILES" %in% colnames(data_orig))) {
      data$SMILES <- data_orig$SMILES # Or the appropriate original SMILES column
  }
  
  compounds_high_cyp_cols_to_select <- c("SMILES", anti_aging_col, cyp_cols, "Num_High_CYP_Flags")
  # Ensure all selected columns exist
  compounds_high_cyp_cols_to_select <- intersect(compounds_high_cyp_cols_to_select, colnames(data))


  compounds_high_cyp <- data[data$Num_High_CYP_Flags > 0, compounds_high_cyp_cols_to_select]

  cat("\nCompounds with high probability (>", cyp_threshold, ") for interacting with one or more CYP enzymes:\n")
  if (nrow(compounds_high_cyp) > 0) {
    if ("SMILES" %in% colnames(compounds_high_cyp)) {
      compounds_high_cyp$SMILES <- sanitize_smiles_for_latex(compounds_high_cyp$SMILES)
    }
    kable(compounds_high_cyp, caption = "Compounds with High CYP Interaction Flags.", escape = FALSE)
  } else {
    cat("No compounds found with high CYP interaction flags above the threshold.\n")
  }
} else {
  cat("No CYP-related columns found or they are not in dataframe. Skipping CYP profile.\n")
}
Profile of CYP Interaction Probabilities.

Profile of CYP Interaction Probabilities. 

## 
## Compounds with high probability (> 0.7 ) for interacting with one or more CYP enzymes:
Compounds with High CYP Interaction Flags.
SMILES Anti_Aging_Prob CYP1A2_Prob1 CYP2C19_Prob1 CYP2C9_Prob1 CYP2D6_Prob1 CYP3A4_Prob1 Num_High_CYP_Flags
8 0.9203264 0.9325538 0.9134086 0.3057511 0.0593774 0.8780224 3
34 0.9120270 0.2669952 0.8276216 0.7600514 0.0616459 0.0584414 2
36 0.7298995 0.6925360 0.8886024 0.4093160 0.0619508 0.0465492 1
40 0.9339135 0.7148754 0.0593912 0.2246064 0.0620323 0.0383697 1
43 0.9068187 0.3091101 0.8622365 0.2279927 0.0616864 0.0584747 1
45 0.6510599 0.2099211 0.5525266 0.7440415 0.0641048 0.3993978 1
53 0.9522983 0.9445683 0.6748180 0.3142129 0.0874351 0.0583887 1
70 0.9182808 0.9216798 0.1105934 0.2611086 0.0575166 0.0584665 1
85 0.9465679 0.3998069 0.7839862 0.5709382 0.0745055 0.2090866 1
88 0.9358224 0.8233892 0.0782701 0.7799495 0.0568602 0.0843939 2
90 0.8675542 0.9144105 0.1819703 0.2400610 0.0463176 0.9125103 2
103 0.8675542 0.9144105 0.1819703 0.2400610 0.0463176 0.9125103 2

Boxplots show CYP interaction probabilities.

Relationship with MTRD

if (mrtd_umol_col %in% colnames(data) && is.numeric(data[[mrtd_umol_col]]) &&
    anti_aging_col %in% colnames(data) && is.numeric(data[[anti_aging_col]])) {
      
  print(ggplot(data, aes_string(x = anti_aging_col, y = mrtd_umol_col)) + # Explicitly print
    geom_point(alpha = 0.6, color = "purple") +
    geom_smooth(method = "lm", se = FALSE, color = "darkred") +
    labs(title = "Anti-Aging Probability vs. MRTD (uMol)",
         x = "Anti-Aging Probability",
         y = paste("MRTD (", sub("\\.", " ", sub("\\.$", "", mrtd_umol_col)),")")) + # Clean name for label
    custom_theme())
} else {
  cat("MRTD (uMol) or Anti-Aging column not found or not numeric. Skipping Anti-Aging vs MRTD plot.\n")
}
Relationship between Anti-Aging Probability and MRTD (uMol).

Relationship between Anti-Aging Probability and MRTD (uMol). 

ames_col_actual_grep <- grep("^AMES.Prob", toxicity_metabolism_cols, value = TRUE, ignore.case = TRUE)[1]
# Ensure ames_col_actual_grep is a valid column name in 'data'
ames_col_actual <- if(!is.na(ames_col_actual_grep) && ames_col_actual_grep %in% colnames(data)) ames_col_actual_grep else NA

if (!is.na(ames_col_actual) && is.numeric(data[[ames_col_actual]]) &&
    mrtd_umol_col %in% colnames(data) && is.numeric(data[[mrtd_umol_col]])) {
  
  print(ggplot(data, aes_string(x = ames_col_actual, y = mrtd_umol_col)) + # Explicitly print
    geom_point(alpha = 0.6, color = "orangered") +
    geom_smooth(method = "lm", se = FALSE, color = "blue") +
    labs(title = paste(sub("\\.", " ", sub("\\.$", "", ames_col_actual)), "vs. MRTD (uMol)"), # Clean name
         x = paste(sub("\\.", " ", sub("\\.$", "", ames_col_actual)), "Probability"),
         y = paste("MRTD (", sub("\\.", " ", sub("\\.$", "", mrtd_umol_col)),")")) +
    custom_theme())
} else {
  cat("\nAMES probability or MRTD (uMol) column not found/numeric, skipping specific toxicity vs MRTD plot.\n")
}
Relationship between Anti-Aging Probability and MRTD (uMol).

Relationship between Anti-Aging Probability and MRTD (uMol). 

if (mrtd_umol_col %in% colnames(data) && is.numeric(data[[mrtd_umol_col]])) {
  valid_prob_cols <- probability_cols[sapply(probability_cols, function(p_col) p_col %in% colnames(data) && is.numeric(data[[p_col]]))]
  if(length(valid_prob_cols) > 0) {
    cor_with_mrtd <- cor(data[, valid_prob_cols, drop=FALSE], data[[mrtd_umol_col]], use = "pairwise.complete.obs")
    cor_with_mrtd_df <- data.frame(Probability_Metric = rownames(cor_with_mrtd), Correlation_with_MRTD_uMol = cor_with_mrtd[,1])
    cor_with_mrtd_df <- cor_with_mrtd_df[order(-abs(cor_with_mrtd_df$Correlation_with_MRTD_uMol)), ]
    
    cat("\nCorrelation of Probabilities with MRTD (", mrtd_umol_col, "):\n")
    kable(cor_with_mrtd_df, caption = "Correlation of Probabilities with MRTD (uMol).", row.names = FALSE)
  } else {
    cat("No valid numeric probability columns found for MRTD correlation.\n")
  }
} else {
  cat("MRTD (uMol) column not found or not numeric. Skipping correlation with MRTD.\n")
}
## 
## Correlation of Probabilities with MRTD ( MRTD..uMol. ):
Correlation of Probabilities with MRTD (uMol).
Probability_Metric Correlation_with_MRTD_uMol
MMP_Prob1 -0.4407394
P.gp_Substrate_Prob1 -0.4321212
TA_Prob1 0.3954930
AIC_Prob1 -0.3855687
P.gp_Inhibitor_Prob1 -0.3811990
CYP2C9_Prob1 -0.3526189
BBB_Prob1 -0.3419256
SCE_Prob1 0.3256494
HLM_Prob1 -0.3130249
DNS_Prob1 -0.3129866
hERG_Prob1 -0.2753722
CS_Prob1 -0.2613587
CYP1A2_Prob1 -0.2541875
CYP2C19_Prob1 -0.2193765
CYP3A4_Prob1 -0.2084049
DILI_Prob1 -0.2049589
MD_Prob1 -0.1904055
EA_Prob1 -0.1886101
GI_Prob1 -0.1078672
Anti_Aging_Prob 0.0900432
CYP2D6_Prob1 -0.0869530
AMES_Prob1 0.0694103
LP_Prob1 0.0537533
Hepato_Prob1 0.0408371

Scatter plots and correlations explore MTRD relationships.

Identifying Promising Compounds

min_anti_aging_prob <- 0.7
max_ames_prob <- 0.3
max_herg_prob <- 0.3
max_dili_prob <- 0.5
min_mrtd_umol <- -4.5 

anti_aging_col_actual_filter <- anti_aging_col 
ames_col_actual_filter <- grep("^AMES.Prob", colnames(data), value = TRUE, ignore.case = TRUE)[1]
herg_col_actual_filter <- grep("^hERG.Prob", colnames(data), value = TRUE, ignore.case = TRUE)[1]
dili_col_actual_filter <- grep("^DILI.Prob", colnames(data), value = TRUE, ignore.case = TRUE)[1]
mrtd_umol_col_actual_filter <- mrtd_umol_col 

required_cols_for_filtering <- c(anti_aging_col_actual_filter, ames_col_actual_filter, herg_col_actual_filter, dili_col_actual_filter, mrtd_umol_col_actual_filter)
# Check they are not NA and exist in colnames(data)
required_cols_for_filtering_valid <- sapply(required_cols_for_filtering, function(x) !is.na(x) && x %in% colnames(data))
required_cols_for_filtering <- required_cols_for_filtering[required_cols_for_filtering_valid]

# Further check if these valid columns are numeric
cols_exist_and_numeric <- sapply(required_cols_for_filtering, function(cn) {
   is.numeric(data[[cn]])
})

if (all(cols_exist_and_numeric) && length(required_cols_for_filtering) == 5) { # Ensure all 5 are valid
  # Ensure SMILES column exists in data for selection
  if (!("SMILES" %in% colnames(data)) && ("SMILES" %in% colnames(data_orig))) {
      data$SMILES <- data_orig$SMILES # Or the appropriate original SMILES column
  }

  promising_candidates <- data %>%
    filter(
      .data[[anti_aging_col_actual_filter]] >= min_anti_aging_prob &
      .data[[ames_col_actual_filter]] <= max_ames_prob &
      .data[[herg_col_actual_filter]] <= max_herg_prob &
      .data[[dili_col_actual_filter]] <= max_dili_prob &
      .data[[mrtd_umol_col_actual_filter]] >= min_mrtd_umol
    ) %>%
    select(any_of(c("SMILES", 
           anti_aging_col_actual_filter, 
           hallmarks_cols[1:min(3, length(hallmarks_cols))], 
           ames_col_actual_filter, herg_col_actual_filter, dili_col_actual_filter, 
           mrtd_umol_col_actual_filter))) %>%
    arrange(desc(.data[[anti_aging_col_actual_filter]]))

  cat("\nPotential Promising Compounds based on Defined Criteria:\n")
  if (nrow(promising_candidates) > 0) {
    if ("SMILES" %in% colnames(promising_candidates)) {
      promising_candidates$SMILES <- sanitize_smiles_for_latex(promising_candidates$SMILES)
    }
    kable(promising_candidates, caption = "Promising Compounds Meeting Selection Criteria.", escape = FALSE)
  } else {
    cat("No compounds met all the specified criteria for promising candidates.\nConsider adjusting thresholds or criteria.\n")
  }
} else {
  cat("\nSkipping identification of promising compounds due to missing, non-numeric, or invalid required columns:\n")
  # print(required_cols_for_filtering[!cols_exist_and_numeric])
  cat("Required columns for filtering were not all valid and numeric.\n")
  cat("Anti-Aging Col:", anti_aging_col_actual_filter, " (Valid & Numeric):", !is.na(anti_aging_col_actual_filter) && anti_aging_col_actual_filter %in% colnames(data) && is.numeric(data[[anti_aging_col_actual_filter]]), "\n")
  cat("AMES Col:", ames_col_actual_filter, " (Valid & Numeric):", !is.na(ames_col_actual_filter) && ames_col_actual_filter %in% colnames(data) && is.numeric(data[[ames_col_actual_filter]]), "\n")
  cat("hERG Col:", herg_col_actual_filter, " (Valid & Numeric):", !is.na(herg_col_actual_filter) && herg_col_actual_filter %in% colnames(data) && is.numeric(data[[herg_col_actual_filter]]), "\n")
  cat("DILI Col:", dili_col_actual_filter, " (Valid & Numeric):", !is.na(dili_col_actual_filter) && dili_col_actual_filter %in% colnames(data) && is.numeric(data[[dili_col_actual_filter]]), "\n")
  cat("MRTD uMol Col:", mrtd_umol_col_actual_filter, " (Valid & Numeric):", !is.na(mrtd_umol_col_actual_filter) && mrtd_umol_col_actual_filter %in% colnames(data) && is.numeric(data[[mrtd_umol_col_actual_filter]]), "\n")
}
## 
## Potential Promising Compounds based on Defined Criteria:
Promising Compounds Meeting Selection Criteria.
SMILES Anti_Aging_Prob AIC_Prob1 CS_Prob1 DNS_Prob1 AMES_Prob1 hERG_Prob1 DILI_Prob1 MRTD..uMol.
0.9606567 0.0006490 0.2034362 0.0039286 0.1204800 0.0594412 0.1322649 37.8386447
0.9495786 0.9859577 0.7616227 0.0114824 0.1426770 0.1134946 0.4244976 7.1163400
0.9446054 0.0164659 0.5620683 0.0398025 0.0335747 0.0584063 0.1793079 0.4631483
0.9406279 0.9928249 0.7415772 0.0252990 0.0118021 0.1402008 0.4668214 7.1163400
0.9405055 0.0240730 0.5064515 0.0123505 0.0106469 0.0689665 0.1417550 0.4631483
0.9404736 0.0311768 0.2085971 0.0064316 0.0369214 0.0357686 0.2666871 143.7607268
0.9382746 0.0782590 0.1542431 0.0049789 0.0273427 0.0399441 0.2351469 143.7607268
0.9301339 0.1079332 0.1822700 0.0076125 0.0044863 0.0520316 0.4187248 37.8386447
0.9295920 0.1179067 0.2875334 0.9805530 0.0674937 0.0384176 0.2963561 14.6569753
0.9295920 0.1179067 0.2875334 0.9805530 0.0674938 0.0384176 0.2963561 14.6569753
0.9235864 0.0303878 0.4370652 0.7006165 0.1460527 0.0570755 0.3112628 14.6569753
0.9212959 0.2429528 0.5325694 0.0355690 0.0058756 0.0778167 0.2893628 143.7607268
0.9203276 0.0339941 0.5515035 0.4313030 0.0562884 0.0554071 0.4157930 14.6569753
0.9173410 0.0013687 0.2106018 0.0049234 0.0007173 0.0366860 0.1743596 14.6569753
0.9160256 0.9962563 0.3491672 0.0016692 0.0862971 0.0799461 0.4669442 7.1163400
0.9127392 0.0557633 0.1630866 0.0035775 0.1254400 0.0504829 0.2995601 143.7607268
0.9110941 0.0172950 0.1774829 0.0141625 0.0233437 0.0717458 0.2086416 143.7607268
0.9110941 0.0172950 0.1774829 0.0141625 0.0233437 0.0717458 0.2086416 143.7607268
0.9098582 0.3543780 0.2370539 0.0964858 0.0062034 0.0629043 0.2629216 0.4631483
0.9069069 0.8692317 0.2914393 0.0146634 0.0070438 0.0707657 0.4678919 37.8386447
0.9065206 0.0341412 0.1985217 0.0062654 0.0796881 0.0420328 0.2085687 143.7607268
0.9041121 0.0044656 0.4256449 0.9521700 0.0829155 0.0301958 0.1725316 14.6569753
0.8909587 0.1037825 0.1533232 0.0157941 0.0286923 0.0534852 0.2939204 143.7607268
0.8860976 0.1298841 0.4931064 0.0151950 0.0193143 0.0576898 0.4279263 14.6569753
0.8820072 0.9254341 0.4331401 0.0018980 0.0368616 0.1089475 0.3992252 7.1163400
0.8740544 0.0250310 0.4150418 0.0117098 0.0002724 0.0731115 0.4904194 143.7607268
0.8740108 0.0170628 0.1947050 0.0015556 0.0000439 0.0438192 0.3600591 143.7607268
0.8690624 0.0218581 0.3847809 0.0123923 0.0004378 0.0759701 0.4824726 143.7607268
0.8675602 0.0159322 0.1808688 0.0016309 0.0336665 0.0377437 0.1922034 143.7607268
0.8675582 0.0104090 0.1560525 0.0065604 0.1791295 0.0435770 0.3559380 37.8386447
0.8675356 0.0469760 0.2333721 0.0033365 0.0022990 0.0595369 0.3350227 143.7607268
0.8675213 0.0149039 0.1791014 0.0013412 0.0001235 0.0350544 0.3186908 143.7607268
0.8675030 0.1906531 0.7086589 0.0888641 0.0106776 0.0753400 0.4017709 143.7607268
0.8674796 0.2518352 0.2211419 0.0079309 0.0029587 0.0853267 0.1555119 143.7607268
0.8674669 0.0673885 0.1724098 0.0018398 0.2561207 0.0593936 0.2268561 143.7607268
0.8674634 0.0094531 0.1458147 0.0047085 0.0126105 0.0336180 0.2520279 14.6569753
0.8674296 0.0793617 0.2277593 0.0047512 0.0008885 0.0422137 0.0878976 3.3764935
0.8634661 0.0707941 0.1232061 0.0195375 0.0024867 0.0559752 0.3830829 143.7607268
0.8601621 0.0015100 0.1796195 0.0020561 0.0038247 0.0409310 0.2049868 143.7607268
0.8269706 0.0164234 0.1609001 0.0076340 0.0001885 0.0386295 0.1531995 14.6569753
0.7944862 0.0966277 0.5920661 0.0138010 0.0009227 0.2311515 0.2995382 7.1163400
0.7857426 0.0997597 0.6192587 0.0147344 0.0001977 0.2195423 0.2521518 7.1163400
0.7592205 0.1638351 0.1449177 0.0044882 0.0040976 0.0646026 0.4647681 37.8386447
0.7040097 0.0027814 0.1907928 0.0054280 0.0139302 0.0643139 0.3060135 143.7607268

This section provides a preliminary filter for promising compounds.

Conclusion and Further Steps

This report provided an initial exploration of the AgeXtend output data.

Key Observations (Example - to be filled based on actual data):

Further Steps could include:

Session Information

sessionInfo()
## R version 4.5.0 (2025-04-11)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.2 LTS
## 
## Matrix products: default
## BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
## 
## locale:
##  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
##  [3] LC_TIME=en_IN.UTF-8        LC_COLLATE=en_US.UTF-8    
##  [5] LC_MONETARY=en_IN.UTF-8    LC_MESSAGES=en_US.UTF-8   
##  [7] LC_PAPER=en_IN.UTF-8       LC_NAME=C                 
##  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
## [11] LC_MEASUREMENT=en_IN.UTF-8 LC_IDENTIFICATION=C       
## 
## time zone: Asia/Kolkata
## tzcode source: system (glibc)
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] tinytex_0.57       factoextra_1.0.7   cluster_2.1.8.1    reshape2_1.4.4    
##  [5] DT_0.33            knitr_1.50         corrplot_0.95      RColorBrewer_1.1-3
##  [9] pheatmap_1.0.12    lubridate_1.9.4    forcats_1.0.0      stringr_1.5.1     
## [13] dplyr_1.1.4        purrr_1.0.4        readr_2.1.5        tidyr_1.3.1       
## [17] tibble_3.2.1       ggplot2_3.5.2      tidyverse_2.0.0   
## 
## loaded via a namespace (and not attached):
##  [1] gtable_0.3.6      xfun_0.52         bslib_0.9.0       htmlwidgets_1.6.4
##  [5] ggrepel_0.9.6     rstatix_0.7.2     lattice_0.22-5    tzdb_0.5.0       
##  [9] vctrs_0.6.5       tools_4.5.0       generics_0.1.4    pkgconfig_2.0.3  
## [13] Matrix_1.7-3      lifecycle_1.0.4   compiler_4.5.0    farver_2.1.2     
## [17] ggsci_3.2.0       carData_3.0-5     htmltools_0.5.8.1 sass_0.4.10      
## [21] yaml_2.3.10       Formula_1.2-5     pillar_1.10.2     car_3.1-3        
## [25] ggpubr_0.6.0      jquerylib_0.1.4   cachem_1.1.0      abind_1.4-8      
## [29] nlme_3.1-168      tidyselect_1.2.1  digest_0.6.37     stringi_1.8.7    
## [33] labeling_0.4.3    splines_4.5.0     fastmap_1.2.0     grid_4.5.0       
## [37] cli_3.6.5         magrittr_2.0.3    broom_1.0.8       withr_3.0.2      
## [41] scales_1.4.0      backports_1.5.0   timechange_0.3.0  rmarkdown_2.29   
## [45] ggsignif_0.6.4    hms_1.1.3         evaluate_1.0.3    mgcv_1.9-1       
## [49] rlang_1.1.6       Rcpp_1.0.14       glue_1.8.0        rstudioapi_0.17.1
## [53] jsonlite_2.0.0    R6_2.6.1          plyr_1.8.9