Ryan's Blog on NMR Software

False Negatives and False Positives are Waiting...

Great post from Derek Lowe from In the Pipeline the other day talking about the dangers of not quality checking those fine-looking starting compounds for your project. Chemistry happens and yes, mistakes do too.

In fact, it appears that Derek has been on a kick as of late referring to personal QC.

I Can Has Ugly Molecules?

Oops.

I thought this would once again be a good opportunity to provide you with a link to a poster Sergey Golotvin presented at ENC 2008 entitled, "Validating the Quality of Large Collections of NMR Spectra Automatically".

Long story short, 15,000 1H NMR Spectra from the Aldrich collection were evaluated in complete automation and the software was able to confirm 88% of the collection as having chemical structures that were consistent with the respective spectra. In addition, 4% were flagged by the software as being inconsistent. A closer, manual look at those 5% revealed that there were indeed some truly wrong structures (or incorrect tautomers) in the collection.

This was evaluating the 1H NMR data only. Using additional 2D experiments, such as HSQC, will likely improve these results.

Just an example of a check an organization can build into their process for additional QC of their registration database for example.

Is it perfect? Absolutely not. There are perhaps a few more false positives in there that the software didn't catch, and of course the software provided some false negatives as well, annoying because presumable someone has to look over them manually only to realize that they were indeed the right structure all along. But at least this doesn't involve manually pining over 15,000 spectra!

We continue to run these datasets, and actually have a consortium consisting of several NMR experts in the industry we call ASCI (Automated Structure Confirmation Initiative) where we are testing and validating this technology in the real pharmaceutical world. Identifying the common areas where false negatives and false positives occur and trying to address them with algorithms.

Will we ever solve all the problems, especially in the world of novel chemistry? Of course not, and for that matter there are some existing problems that appear to be too hard to solve.

But that being said, what is the acceptable limit of false positives and false negatives for automated verification by software for the verification of  registered compounds in a library?

Interested in hearing your thoughts.

More News and Catching up Around the Blogosphere

In my last post I provided my excuse for not blogging much as of late.

In addition to that news, I have some other news to report. Over the past 5-6 weeks I have transitioned to a new role and I am now the NMR Product Marketing Manager for ACD/Labs. This role comes with many responsibilities, but most importantly, I am now considered the coordinator and primary point of contact for the strategic business directions of the NMR Product Line. I am thrilled to take on this added responsibility and am looking forward to working even closer with our NMR Development and NMR Technical Teams.

So perhaps now your input on this blog will have an even larger impact :)

Enough about me, during my absence there have been some pretty intriguing things being written in the blogs I frequent:

In Oops, Derek Lowe talks about an unfortunate and rather common issue in the industry. Specifically:

"But there are more of these things waiting out there – in every large compound collection, in every catalog, in every collection of data are mistakes. Watch for them."

I'll definitely be referring back to this one for a future blog post.

Glenn Facey and Arvin Moser continue to provide tips and insight on the acquisition, processing, and interpretation aspects of NMR on their respective blogs.

ChemSpider announced the ability to perform NMR predictions via ChemSpider only to later announce the removal of the service. Tony then applauded the NMRDB.org development staff for acting quickly on the issues identified in a public forum.

Weiwei TIAN, has found one of the features in 1D NMR Assistant and/or 1D NMR Processor to be very useful.

Speaking of 1D NMR Assistant, ASDI shares their motivation behind their decision to deploy this software to their chemists.

Happy reading!

Where's Ryan?

It's been over 2 weeks since my last post and I apologize for my absence in case anyone has missed me :)

I am however happy to share that there is a good reason I have not been blogging.

Last Sunday, my wife gave birth to a beautiful baby boy and I have spent the last 2 weeks bonding with the little guy.

Ezra Donovan Sasaki

Again, my apologies for the personal note, but if you can't tell I'm one proud papa!

SMASH 2008-Session on Computer Aided Structure Structure Identification

Speaking of the CASE review article in Progress in NMR, I thought I would take the opportunity to mention the upcoming session on Computer Aided Structure Identification that will take place at the SMASH 2008 Conference in Santa Fe New Mexico in September.

One of the presenters in the session will be none other than Mikhail E. Elyashberg who has made tremendous contributions to this area. In his presentation he plans to lay out the 40 year history of CASE and I am sure it will be a terrific presentation.

8 years ago Mikhail was one member of a team of 4 who received the State Award of the Russian Federation for their pioneering work into the automated analysis of complex compounds. It is great to see Mikhail continue to work, publish, and present on this topic.

Rich Beger from the National Center for Toxicological Research and Richard Newmark will also be presenting in this session.

Not to mention that Gary Martin (another Co-author of the CASE article in Progress in NMR) is slated to speak in the New Experimental NMR Techniques session.

So there's a number of reasons for you to get away from the lab for a few days in September and attend SMASH 2008.

Looking for a Great Weekend Read?

In fact, to borrow a phrase from a colleague, this might be the defacto article on Computer Assisted Structure Elucidation (CASE) for the next decade!

This article written by Mikhail Elyashberg, Antony Williams, and Gary Martin spans across two issues of the review journal, Progress in Nuclear Magnetic Resonance Spectroscopy. 

This article entitled, entitled, "Computer-Assisted Structure Verification and Elucidation Tools in NMR-based Structure Elucidation" is available online and you can review a preview of the content at:

http://dx.doi.org/10.1016/j.pnmrs.2007.04.003

This is a very important and comprehensive review of modern expert CASE systems over many years. It includes specific examples of complex natural product structures that have been automatically elucidated using such systems.

I thank the authors of this publication for their contributions in this area, and the efforts they have now put forth to communicate this story to the scientific community.

Please obtain a copy for yourself, I can promise that it is a very informational and intriguing read for those of you who do NMR regularly.

Time Flies...1 Year Anniversary of Ryan's Blog on NMR Software

I can't believe it has been one full year since my very first blog post. 73 posts 90 comments, >500 subscribers later...here we are.

I hope that over the past year I have accomplished what I set out to do: to provide an updateable resource for people interested in NMR software, more specifically ACD/Labs software.

Instead of attempting to write a meaningful post that would appear fitting, I thought I would reflect back on my favorite 5 (OK, I couldn't resist I had to go with 7!) posts of my rookie year as a blogger:

Here they are, IMO:

7. The Truths and Myths of ACD/Structure Elucidator
6. The Price of NMR Software
5. The Curse of Knowledge
4. Where is the Quality? Assumption vs. Assurance?
3. Fringe Benefits and Knowledge Management
2. NMRShiftDB, ACD/Labs, and Modgraph (This prompted quite a lot of dialog around the web, there's a lot of reading here)
1. The Purgatory Database

A big thank you to all those who have subscribed to my blog and those who have offered comments, emails, etc. that have inspired me to continue on with this endeavor. A blog is only as good as it's followers.

A big thanks to Antony Williams, Gary Martin, Phil Keyes, Rich Apodaca, Wolfgang Robien, Malcolm Beckett, Jeff Seymour, Jean Claude Bradley, Egon Willighagen,  Mark Mowery, Christoph Steinbeck, and Stefan Kuhn for their inspiration, insights, opinions, comments, and overall spreading of the word through the blogosphere and beyond.

I would like to thank Daria Thorp, Antony Williams, Brent Lefebvre, Mark Bayliss, Arvin Moser, Andrew Anderson, and Patrick Wheeler for their internal support.

And of course our NMR Development Project Leaders Sergey Golotovin, Kirill Blinov, and Mikhail Kvasha without of course this blog would not exist.

Here's to another year of NMR Software blogging!

Automated Structure Verification by 1H NMR Only

I've blogged several times about the progress and applications of automated structure verification with the help of ACD/Labs software.

There are really two main approaches right now:

1. Combined verification which includes automatically verifying the correspondence between a proposed chemical structure and the 1D ¹H and 2D HSQC spectra. I've blogged about the publications and applications of this method previously.
2. ¹H only verification. Of course the first approach is preferable from an accuracy standpoint as the additional information gained from the HSQC spectrum increases the selectivity and specificity of the results.

However, I always get questions on how well we perform on ¹H NMR only because in some organizations and environments it is simply not feasible to always run an HSQC in tandem with a routine ¹H NMR analysis.

As mentioned in previous posts, we've already conducted a validation study on this approach and it was published in a 2006 article in MRC. We've continued to investigate and validate this approach and we recently presented our latest results using 1D ¹H NMR data only at ENC 2008.

The poster highlighted a study on the automatic evaluation of over 15,000 Aldrich compounds and spectra from Aldrich NMR Spectra Database.

The results of this study revealed that the software was able to confirm 88%of all spectra as consistent and flagged less than 5% as inconsistent.

One of the more interesting discoveries in this study was that it revealed some truly wrong structures in the Aldrich NMR database.

More information on these as well as shortcomings in prediction, processing, and analysis are provided in the poster that can be downloaded:

http://www.acdlabs.com/download/publ/2008/enc08_aldrich.pdf

Do You Run 15N, 19F, and 31P Experiments?

As I am closely approaching my one year anniversary of this blog (how time flies!), Arvin's post entitled, "How Do I Know if my Unknown Contains a Fluorine Atom",  reminded me an X-nuclei related post from almost a year ago.

In that post I highlighted a best practice document written by Gary Martin for acquiring ¹⁵N-¹H heteronuclear shift correlation data where he highlights an interesting application for ¹⁵N NMR prediction.

If you think that only benefits of a ¹⁵N NMR Predictor is for structure verification or validation, Gary is providing you with some additional tips and suggestions.

Check out his document here.

I want to personally thank Gary for his incredible contributions to the NMR community as well as for his guidance, collaboration, and contributions to the ongoing development and growth of ACD/Labs software.

Love/Hate Continued...

To follow my post from yesterday, Rich made a very interesting comment that I was hoping to address in today's post.

He suggests that perhaps a NMR Lite would be a good approach. I gave reason as to why we didn't take that approach and strip away existing features from 1D NMR Processor when developing the processing component in ACD/1D NMR Assistant yesterday.

But Rich makes a very good point in his comments:

For everyone else, it's all too easy to not even look at a piece of software who's fundamental purpose is obscured by layer upon layer of expert features. They just tune out and the only ones who end up using the software and giving feedback are the power users - which unfortunately reinforces the misconception.

For me, the key is to make sure that all these features don't get in the way of the primary reason someone is using the software. I don't have a problem with a wealth of features as long as it does not interfere with the main workflow.

Cathy Sierra is a blogger that I dearly miss for her daily insight on how to "create passionate users". Here's one of her takes. Specifically I like:

One of the themes I heard over and over at ETech and SXSW (Jason Fried, Craig Newmark, and others) was the developer mantra of "get out of the way." In other words, build the thing so that it stays the hell out of the way and lets the user get on with what they really want to do.

 

So as an adaptation of Rich's comment. Make the 20% REALLY clear, and hide the other 80% but still offer it.

I'm not sure it's the perfect solution, but I think it's a different approach.

For example, here's the first thing a user will see when they open a raw data file in 1D NMR Processor:

On the other hand, here is ACD/1D NMR Assistant:

Of course you don't want to create an environment where a user is drilling through menus looking for useful features, but we do provide users with the ability to hide or show toolbar buttons and action buttons on the interface so they can choose the interface most appropriate to them which can be altered as they get more comfortable with the software.

What do you think Rich?

Anyone else have an opinion they want to share in the comments section?

Avoiding the Love/Hate Relationship in Software

I had a conversation with Geoff, one of my ACD/Labs colleagues just yesterday.

He provided me with a great quote from a person he was talking to about software and usability.

He said:

"User-Friendly...hrmph...what that means to me is: love it for the first week, hate it forever!"

I think that's a great quote.

I think usability is incredibly important and should take high priority in the development of a software package.

However, while it is important we can't forget the term USE in "Ease of Use"

I find a lot of software packages out there that are very pretty, very intuitive, and very easy to use. They make a great first impression. You can use it for few days and fall in love. However, often times what happens is that after a week you decide you want to do more sophisticated things with the software but you can't. All of a sudden that love, turns to frustration, and sometimes ends up in hate.

I think this is one of the more difficult aspects of software development. I think most Product Managers and developers want their products to be easy to use. But you can't go overboard and oversimplify.

We went through the simplification process with the development of ACD/1D NMR Assistant. In the very early stages, we thought it might be a good idea to create a 1D NMR Processor Lite for chemists. We knew that 1D NMR Assistant would have the same NMR processing component as ACD/1D NMR Processor. But we also came to the conclusion that the interface and workflow in a duplicated form would not be appropriate for a new market, after all ACD/1D NMR Processor was developed for years with the NMR Spectroscopist in mind as they were the ones using it, providing feedback, and NFRs (New Feature Requests).

So while ACD/1D NMR Processor contains many features we don't anticipate most chemists will make use of (a quantitation tool, macro capabilities, group macro capabilities, intelligent bucketing, multiple baseline and phase correction algorithms, arithmetic, peak fitting, etc.) we decided to leave EVERYTHING in there. Perhaps 1 in 50 will become frustrated when their software can't execute a particular function for them, we're hoping we've avoided this issue by including all the features available in 1D NMR Processor within the Assistant package.