Berkeley defines high dimensional analysis generally as:

High-dimensional statistics focuses on data sets in which the number of features is of comparable size, or larger than the number of observations[1]

Basically what high dimensional data analysis means is that you have a data-set with multiple variables (or dimensions) that need to be statistically analyzed to make sense.

For example, an analysis of a data set from a clinical study involving 100 patients with 10 different timepoints of blood pressure measurement values, and the profiling of blood cells recorded with multiple parameters, such as gender and age etc., all together is high dimensional analysis.

In flow cytometry, high dimensions mean high numbers of parameters are used to visualize molecules expressed on a hematologic cell, and its analysis has come from an improvement in the base technology. This improvement has allowed a much higher number of parameters to be analysed per cell. Anything with over 50 parameters is considered to be mass cytometry or CyTOF.

High Dimensional Data Provides Flow Cytometry Opportunities

Developed in 2009, CyTOF is a new sector of flow cytometry that could question over 50 parameters per cell. The excitement about this new development was evident by the number of dedicated CyTOF departments that were established worldwide.

This increase in parameters was made possible by an instrument that can theoretically detect nearly 100 available heavy metals, compared to the conventional fluorescent flow cytometry output, which allowed just 18 parameters at maximum. The fluorescent flow cytometry data has traditionally been analysed by manual gating on bivariate plots, which is notoriously time-consuming and can result in bias and undesirable variations between researchers analysing the data set.

The manual gating strategy used to analyze flow cytometry data has become basically impossible for high dimensional data sets, because of the complexity which is beyond the human brain can interpret. Flow cytometry high dimensional analysis software uses automation and powerful data engines to save time and provide detailed results, removing this bias.

The biotechnology industry has had to find a way to incorporate multi-parametric characteristics of these new data sets, which result from high-throughput methods.

Together with mass cytometry and new computational data analysis strategies, scientists have been able to draw more detailed conclusions about cellular heterogeneity and characteristics from their single-cell data and expand the range of diagnostic and therapeutic applications.[2]

Software Is Changing For High Dimensional Data Analysis

Companies providing software for flow cytometry data analysis have had to run fast to keep up with this newfound technology. There are an array of products online, all of which have different features to work with this data and provide different user experiences.

There are several important features when looking for software to analyse high-dimensional data.

  • A range of statistical analysis options allows you to customise how you view data from your high number of parameters.

  • A large amount of storage space ensures you have enough for larger datasets.

  • A range of data visualisation tools makes analysing the results as easy as possible for such a large amount of data.

Tercen - a High Dimensional Data Analysis Software Solution

At Tercen, we provide all the features listed above to improve your high dimensional analysis experience. The platform can be as complex or easy to use as you like, as you add in extra options and features.

Our software operates with a drag and drop interface, making everything very visual and simple to move around. We offer multi-variate plots and heatmaps, as well as the usual bi-variate plots, to enable you to customise your data analysis as far as you need to.

In terms of statistics, we offer clustering and dimension reduction and differential statistics, and meta-annotation. We offer integration for single-cell RNAseq and any other molecular readouts you might need to analyse.

You can use R plugins, but we also offer python plugins and central exchange to enable you to integrate your data from a range of different sources without learning how to code

Reports can be customised and made online and exported to a PDF with the click of a button.

One of the most unique things about Tercen is our focus on the scientific community. You can share and work on data collaboratively, get feedback on your projects, and share your results with the rest of the Tercen community.

Find out more and sign up below:

[1] Statistics at UC Berkeley. [2021]

[2] European Pharmaceutical Review. [2012]

Did this answer your question?