Brute force analysis.

Modified on Wed, 13 Sep, 2023 at 2:40 PM

TABLE OF CONTENTS


Introduction

In computer science, brute force is a very general problem-solving technique and algorithmic paradigm that consists of systematically enumerating all possible candidates for the solution and checking whether each candidate satisfies the problem’s statement. Think about checking the gene expression of a gene in a certain dataset. Is that gene expression part of a broader pattern that identifies a phenotype, or is it an isolated event?

In datasets with a high number of observations (like single cell RNA-seq datasets), one of the main objectives is to identify clusters in the data that correspond to distinct phenotypes. When the biological differences between these clusters are stark, then it is easy to understand which cluster belongs to a specific cell phenotype. The problem arises when the differences between these clusters are not stark,but is a continuum in which the clusters blend into each other. In this situation, assessing the number of clusters (with the Dimensionality reduction analysis plot) present in the data becomes problematic. Even though these clusters blend into each other, they should have a distinc expression signature that differentiate them (otherwise they would be the same cluster).

To find this distinct expression signature, or these patterns, the UniApp plots all the features available in the data using the current Dimensionality reduction plot,and arranges them in an easy-to-use table that can be consulted and scrolled through very quickly. By consulting and exploring this table, you can guess the number of clusters that are present in your data.

In short, the brute force plotting analysis helps you in finding the true number of clusters in your data, which you will then use in the Clustering analysis.

The input for the bruteforce analysis is the currently selected dimensionality reduction analysis step. 


1 Creating a plot 

 


As a first step of the analysis, a plot must be created by clicking on the create plot icon in your analysis track. This will lead to a section where the analysis of interest can be selected.



In order to ensure efficient organisation, a name and description must be assigned to the analysis under the appropriate fields. Subsequently under "Choose algorithm to run your analysis" Bruteforce must be selected.


2 Selecting data


In the field "Choose track element", the input analysis can be selected. In the cell selection tab you can choose the observations to use as input. For more information see the section on Cell/sample selection. Note that subsetting at the pretreatment step is a "hard" subset meaning that excluded cells/samples at this step will not be present in the downstream steps.



3 Setting parameters


In the Set parameters field you will be able to define how to perform the bruteforce analysis.

  • Parameters: sets dot size. Also toggles test run. With test run you can check if the dot and plot size is suitable before plotting all of the features.
  • Advanced: plot size and quality.

1.1 Bruteforce customization

In the Parameters and Advanced tabs you can customize the appearance your bruteforce analysis result panel. 


1.1.1 Parameters



In the Parameters tab you can select the dot size size on the generated plots. You can perform a test run to generate just a single plot in the results panel to determine if selected customization settings are suitable for you. 


1.1.2 Advanced


In the Advanced tab you can selected the individual plot width, height and DPI (dots per inch).

4 Performing the brute force analysis


When the parameters are all set-up, you can click on the Run button to generate all the plots. This could take quite some time, depending on the size of the data.

As soon as the generation of the plot is computed, a panel of plots will appear, where each plot in the panel is color-coded for one of the features in your data matrix. Now you can start exploring the results of the bruteforce analysis. Be sure to look out for patterns that reflect genuine cell phenotypes. You can then use this information to get a sense for the correct resolution for the clustering analysis. 

The plots are based on the currently available Dimensionality reduction results. If no result is available, you will not be able to perform the brute force analysis.


5Bruteforce analysis visualization 



In the visualization tab you can adjust the number of rows, columns and the size of the individual plots in the bruteforce analysis results panel. 

Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select at least one of the reasons
CAPTCHA verification is required.

Feedback sent

We appreciate your effort and will try to fix the article