A major challenge of analyzing time-series experiments is that they often differ in the length of recovery from synchrony and the cell cycle period, so measurements from different experiments cannot readily be compared or analyzed in the aggregate without alignment. Clocks lifeline alignment is a method for phase-specific and biologically relevant comparisons between experiments and for aggregating multiple replicate experiments, both of which were previously difficult or impossible. Previously, phase-specific and biologically relevant comparisons were made by tracking landmark events.
However, using these ad hoc methods, subtle but important differences remain undetected. Comparison between time-series experiments is particularly difficult between experiments with significant differences in timing, such as in mutant populations or growth conditions that affect the synchrony recovery time or the cell cycle period. Clocks lifeline alignment allows for phase specific comparison in these cases, We have used Clocks to align time series transcriptomic and proteomic data, which allowed for the direct comparison between the dynamics of the mRNA and the corresponding protein.
Furthermore, we align time-series experiments across different species, providing critical insights into cell cycle evolutionary changes. To begin, download the Clocks alignment repo by entering the command into the terminal. Create a conda environment using the condarec.
yml file by entering the command into the terminal in the folder where the Clocks alignment repo was cloned. Double click the condaV2023. jar file located in the Clocks folder in the Clocks alignment repo, and wait for a graphical user interface to open.
The screen allows input options for the Clocks run and displays the results once run. Input the experimental conditions by specifying the temperature in Celsius using the text box labeled temperature, and specify the synchronization method using the dropdown menu, synchro.method. To input the settings for the budding data, choose the option bud for budding yeast from the model type dropdown menu.
Import the data by uploading a file using the select file button or by typing it into the text input boxes of the data import panel. The first column specifies the time points and the other two columns specify the budding data. To input the settings for flow cytometric data, select the flow section from the model type dropdown menu.
Import the data using the data import panel and click select file, then select the time points in the times for fitting box, for which a flow cytometric Clocks fit should be plotted. Once all the inputs have been selected for either budding or flow cytometry, click the apply button and then the sample button at the top of the screen. View the budding curve or flow cytometry plots in the predicted fits tab.
Obtain the Clocks parameters from the fit by selecting the posterior parameters tab. The resulting table has each row consisting of the parameter, with the final row being the posterior. The columns consist of the predicted parameter for the mean, 2.5%lower confidence interval, 97.5%upper confidence interval, and acceptance rate.
The Clocks budding curves were properly fit, as demonstrated by the data points overlaying the corresponding fit curve with a small 95%confidence band. With flow cytometry cell cycle phase data, Clocks produces a Clocks fit for each selected time point. To begin, activate the conda environment by entering the command into the terminal.
Use the command to open an interactive Python notebook. Create a new Python notebook in the desired folder, providing an appropriate name, then import the Python file containing the alignment functions by running the command in the first cell. If using budding data as the cell cycle phase data, import a data frame containing the percent butted at each time point by running the command in a new cell, then align the budding data to a lifeline point timescale by entering the function into a new cell.
Next, import the experimental data frame into the notebook by running the command in a new cell. Align the experimental data to a lifeline time point scale by entering the function into a new cell. then enter the command into a new cell to download the lifeline aligned data set.
The budding data collected from the condition 2 RNA-seq showed the percent butted over time for both the unaligned timescale in minutes and the aligned timescale in lifeline points. The flow cytometry data collected for the condition 2 dataset plotted for a selected time points, showed that the data matched the phase determined by the alignment. Plot the budding curves before alignment using the Python utility function by entering the command into a new cell.
Next, plot the budding curves after alignment using the Python utility function. Use the provided plot line graph comparison in the Python utilities vial to perform line graph comparisons on the original, aligned, or aligned and interpolated data frame by typing the command into a new cell. Import a CSV or TSV gene list file into the notebook using the command in a new cell.
Next, use the provided function plot heat map comparison in the Python utilities file to perform a heat map comparison on the aligned, interpolated, and phase-aligned data frame by typing the command into a new cell. A comparison of the aligned and unaligned transcriptomatic data showed that before alignment, the first peak expression of the microarray experiments appeared aligned with the second peak of the RNA-seq experiment. However, after alignment, the first cell cycle peaks of each dataset are appropriately aligned.
Comparison of the cell cycle phase data across experiments with varying periods exhibited visible period differences in the unaligned budding curves, whereas Clocks alignment made the three curves look remarkably similar, making comparisons of experimental data possible. The cell cycle phase data for each comparable lifeline point was not identical between the two conditions. Comparison of the transcriptomic data across experiments with varying periods showed that before alignment, the transcript dynamics of CDC 20 were non-overlapping, but after alignment, the peaks occurred on the same cell cycle phase, but the shapes of the curves were different.
The genes were plotted as heat maps in the same order for all three conditions for both unaligned and aligned.
