I have been assigned a task and I am a little bit out of my comfort zone. I am looking for software recommendations for HPLC statistical analysis that can incorporate several factors/indicators such as retention time or peak area.
We have been given a data set of numerous samples that each have two different HPLC traces at 280 nm and 210 nm which have peaks at different retention times. The desired result is to perform hierarchical clustering such that the analysis will take into account both wavelengths when comparing between samples.
Additionally, if anyone knows of any reliable pipelines for handling these sorts of analysis that would be greatly appreciated.
The data sets you have should be sorted based on WAVELENGTH first, then sorted by Retention Times and perhaps peak areas. Most chromatography data systems (CDS) already provide output in this format. That output could be an Excel or CSV format for import.
- Regarding Hierarchical Cluster Analysis software. If you do a search on the web, you will find lots of free and also 'paid' software packages available for download. Example: https://www.ncss.com/software/ncss/clustering-in-ncss/
Note: If the chromatography Rt & wavelength(s) chosen (not by you, but by the person who ran the original methods and collected the data) were not optimized for the samples, any conclusions drawn from statistical analysis would have no practical validity. Statistical analysis can be run on most any data set, but that does not mean the data or conclusions drawn from it are useful or even accurate. *The data gathered through chromatographic methods can be very subjective (it is not like data gathered from other instruments such as a balance or pH meter which are more objective measurements). Please be very cautious when selecting data and forming any conclusions.
I am going to 'muddy' your water!
In general the chromatographic data software (CDS) of the HPLC complies with 21 CFR part 11. In addition it contains common equations (like resolution).
Similarly, the LIMS (Laboratory Information Management System) complies with 21 CFR part 11, but can contain unverified equations. Thus, many places use SAP, SAAS....
The data sets you have should be sorted based on WAVELENGTH first, then sorted by Retention Times and perhaps peak areas. Most chromatography data systems (CDS) already provide output in this format. That output could be an Excel or CSV format for import.
- Regarding Hierarchical Cluster Analysis software. If you do a search on the web, you will find lots of free and also 'paid' software packages available for download. Example: https://www.ncss.com/software/ncss/clustering-in-ncss/
Note: If the chromatography Rt & wavelength(s) chosen (not by you, but by the person who ran the original methods and collected the data) were not optimized for the samples, any conclusions drawn from statistical analysis would have no practical validity. Statistical analysis can be run on most any data set, but that does not mean the data or conclusions drawn from it are useful or even accurate. *The data gathered through chromatographic methods can be very subjective (it is not like data gathered from other instruments such as a balance or pH meter which are more objective measurements). Please be very cautious when selecting data and forming any conclusions.
- Badges
- Science method