Understanding Contamination Between Cells in a Grid

In this article, we’ll delve into the process of identifying contamination between cells in a grid. The task involves analyzing weight measurements from each cell and determining whether there’s evidence of cross-contamination.

Background and Context

The scenario presented involves a machine that drops microscopic particles into cells within a plate containing 96 cells (8x12 grid). After the machine is finished, the weight of each cell is measured. The goal is to identify potential cases of cross contamination by combining the weight information with spatial data from the grid.

Step-by-Step Process

To tackle this problem, we’ll follow these steps:

1. Convert the linear data into a useful matrix in R
2. Calculate the median weight of all cells
3. Define the threshold for identifying potential cases of cross contamination
4. Implement the script using R and its built-in functions

Step 1: Converting Data into a Matrix

To efficiently analyze the data, it’s essential to convert it into a matrix that allows us to easily access neighboring cells.

library(data.table)
DT <- as.data.table(my.data)

In this step, we use the as.data.table function from R’s data.table package to convert our linear data into a more suitable format for analysis.

Step 2: Calculating Median Weight

We need to calculate the median weight of all cells in order to determine the threshold for identifying potential cases of cross contamination. We can use the built-in median function from R’s base library to achieve this.

median.weight <- DT[, median(Weight)]

Step 3: Defining Threshold

Next, we define the threshold for identifying potential cases of cross contamination. In this scenario, any cell with a weight greater than or equal to 1.5 times the median weight should be checked against its neighbors.

# Define threshold for contamination check
contamination.threshold <- median.weight * 1.5

Step 4: Implementing Script

Now that we have all the necessary components in place, let’s implement our script using R and its built-in functions. We’ll create a new column called Contamination to store the results of our analysis.

# Create Contamination column based on neighbors' weights
DT[, 
    Contamination := ifelse(
      Weight >= contamination.threshold & 
      ((.I %% 8 != 0 & shift(Weight, n=1, type="lead") < 1) | # not in last column, check next value
        (.I %% 8 != 1 & shift(Weight, n=1, type="lag") < 1) | # not in first column, check previous value
        (.I<88 & shift(Weight, n=8, type="lead") < 1) | 
        (.I>8 & shift(Weight, n=8, type="lag") < 1)), 
      TRUE,
      FALSE
    )
]

In this final step, we use the ifelse function to create a new column called Contamination. This column will contain TRUE if the current cell’s weight is above the contamination threshold and its neighbors have weights below 1. Otherwise, it will be empty (NA).

The Result

With our script in place, we can now analyze the data and identify potential cases of cross contamination between cells in the grid.

Here is a sample output:

In this sample output, we can see that cell F3 has been identified as potentially contaminated due to its high weight (above the contamination threshold) and neighbor F2, which is nearly empty.

By following these steps and implementing our script using R, we’ve successfully analyzed the data and identified potential cases of cross contamination between cells in the grid.

Last modified on 2023-05-18