QbD Software (ChromSword, ACD, S-Matrix)

[dwcook43]

We're looking to invest in some QbD focused software for automated method development activities. I'm focused on ChromSword, ACD/Labs, and S-Matrix as they seem to be compatible with our Agilent heavy lab.

Best I can tell is S-Matrix is focused on DoE, where as the other two seem to integrate some chromatographic theory (but I may be wrong). But I'm struggling to come up with meaningful comparisons between the three.

Does anyone out here have any experience with one or more of these? Any insight would be helpful

Thanks in advanced!

[Multidimensional]

A few comments. Yes, they are all very different packages. ALL require that YOU first use your own experience (should be > 15 years full time bench chromatographer) to set up the initial screening methods needed to generate data. The software is "dumb" so still relies on an expert to create the initial, well chosen screening methods, and also interpret the results (as the software can not tell which peaks are sample(s) or even real (humans still do that). If you give the software initial methods which are poorly thought out using columns which are not ideal for the analysis, then the software will give you equally bad results. While these packages may be sold as "automated method development" software, they are not. They simply use some basic fundamentals of chromatography (that a trained chromatographer would and should know) to suggest when to change composition, flow. temperature etc in an attempt to improve the resolution. If you have ever used "DryLAB", then you have used a much better version of the software. DryLab follows the basics, in an ideal way, but is a good teaching tool. It does a great job of showing how temperature, flow rate, composition etc may change the theoretical separation. You can do a lot of trial and error runs, simulations with it, as well as use real data from analysis runs (This is what the more expensive pkgs do too). Any well trained chromatographer can out do these software packages you mention. A novice or intermediate user does not have the method development experience or training yet to direct the software in a rational way or help it make decisions where to 'go next'. THE SOFTWARE is 'dumb', fully relying on you for initial column and mobile phase choices (this is why it fails). Many of the suggested steps are not practical. Additionally, for the software to be useful, you will also need a fully automated and integrated column selection system as part of the HPLC. COLUMN selection is the more important variable in HPLC method development, followed by mobile phase composition.

• If you really want to improve your ability to develop HPLC methods, then hiring a more experienced chromatographer, not using awkward, difficult to program software is the solution. Software is not the answer. Training is.

Considering the purchase of any advertised "Automated HPLC Method Development" software product with the goal of increasing your productivity will fail. The sales people who sell these products want you to believe their product is the answer. The less experience you have in HPLC, the more likely you are to believe them. Once you have more advanced level training and experience, you will understand this. Don't waste your money on it. Spend it on training or hiring more skilled users.

Lastly, from our lab's experience using ACD and ChromSword (the two most popular), Chromwsword is by far the most difficult to learn and use software. Many of our clients have the most difficulty setting up and using it.

[dwcook43]

I agree with you that software is no replacement for good chromatographic theory and experience, but I have to strongly disagree with you that one needs >15 years at the bench in order to develop good chromatographic methods.

The view we are taking is that this software can be a tool in our arsenal to expedite the scouting process, particularly since we are a small analytical group with many duties other than method dev. We plan on either outfitting an existing system with a larger capacity column compartment, or potentially adding a new (mostly) dedicated system to our lab.

I do appreciate your insight at the end of your post regarding the difficulty in implementing ChromSword.

[Multidimensional]

"but I have to strongly disagree with you that one needs >15 years at the bench in order to develop good chromatographic methods"

- I will agree that it is very hard to suggest a time interval for the needed training and experience, but we never come across users with less than ten years of experience who know the basics. Most need experience at different labs, doing different things, while obtaining training to be proficient at MD. My boss, who has over 30-years of professionally teaching chromatography to scientists worldwide and has designed and developed several of the instruments used in the industry (esp many of the automated chromatography method development screening systems used in all of the top labs worldwide) has found that no one really needs the "special" software sold for this purpose. Again, DryLab can teach you the basics. If they have obtained the proper training and acquired the needed knowledge, then they already know how to progress in an efficient manner through the method development process (99% do not, so software will not help them). The "special" software does not add a needed tool to their toolbox. However, an automated column selection system (correctly set up) does add a tool to their toolbox because no human can efficiently screen a range of columns on a system as well as a machine can.

"The view we are taking is that this software can be a tool in our arsenal to expedite the scouting process";

Properly configured by an experienced user, an automated COLUMN screening system using stock software (CDS) can quickly narrow down the best column(s) and mobile phase composition far better than any person using an LC alone can. This we know by actual experience having developed thousands of novel LC methods for many different application classes since the late 1980's. The so-called method development "software" mentioned always relies on the knowledge and training of the operator for directing the search and interpreting the results (which is why it adds very little value).

[JoNeu]

Hello,
i have about 8 years experience in working with Fusion and DryLab for pharmaceutical method development. I also had a trial period with ACD.
All will not help you in evaluating scouting experiments. They shine during optimization steps. You need to find a suitable column chemistry and additive combination first.
When we are talking about purity methods for pharmaceuticals that need to separate 20 or compounds in a robust way you will not get around one of those softwares. Even very well experienced chromatographers will not be able to optimize such methods in a time effective way. When working in a pharmaceutical enviorement current guidelines (ICH Q14 and USP 1220) also need to be respected. These require the manufacturer to show QbD data on robustness of the final method. Software is necessary to visualize the data. Anyway, you need to pair a trained chromatographer with these packages or you will not see any benefit.