Advancements in NMR Predictions- Neural Network vs. HOSE ... · PDF fileAdvancements in NMR Predictions-Neural Network vs. HOSE Code Algorithms ... Prediction Accuracy ... (ppm)-40-20

Advanced Chemistry Development, Inc. (ACD/Labs)

Advancements in NMR Predictions-

Neural Network vs. HOSE Code Algorithms

Brent LefebvreNMR Product Manager

ACD/Labs’ ENC User’s MeetingApril 22, 2007

2

Why Neural Networks?

The Neural Network algorithm offers a very specific advantage

Speed of calculation is hundreds of times fasterThis enables prediction on-the-fly

For Structure Elucidator, a key feature

3

Why Neural Networks?

A fresh approach for ACD/Labs to shift predictionWe are always researching new ways to improve our software

Also see our poster (#150) on our new increments scheme

4

Realization

The Neural Network algorithm was outperforming our version 9 HOSE code!Steps were then taken to migrate this algorithm out of Structure Elucidator and into the ACD/CNMR Predictor

5

Implementation

6

Neural Network Algorithm

7

Implementation

Training the Neural NetEntire database from version 9 usedAdditional database of 187,000 shifts used for accuracy testing

8

Neural Network Approach

How does this neural net implementation compare to others in the industry?What is unique about it?Does this make it better or worse?

9

Neural Network Approach

Our research brought us to some new conclusionsSome implementation details differed from previous industry attempts

10

Neural Network Approach

We found that:Characteristics of the Neural Net were NOT the most important factorStructure encoding scheme was most importantSize and accuracy of training set is key

Our huge quality checked database gave us a tremendous advantage

11

Using the Neural Network Predictions

How are they accessed in the software?

12

Using the Neural Network Predictions

13

Using the Neural Network Predictions

14

Limitations of the Neural Network Predictions

Predictions are a black boxNo calculation protocol as for HOSE code

Training of predictions could be possible

Does not outperform HOSE code training

15

Statistics

How do NN compare to old and new HOSE code?When should I use NN?What is the new performance?

16

Prediction Accuracy

We calculate our prediction accuracy for HOSE code the same way every year

A “Leave-one-out” analysis of our entire database (2 million chemical shifts)

This allows us to compare year on year improvementA TRUE analysis of how accurate the predictors are

17

L-O-O Analysis

Database: W:\CNMR.1000\STATISTICS\CNMR105.INTChemical Shifts : 10.5 (1982234 pts)

280260240220200180160140120100806040200-20-40-60Chemical Shifts : Value (ppm)

-40

-20

0

20

40

60

80

100

120

140

160

180

200

220

240

260

280

Database: D:\TEMP\FROM48\CNMR.800\CNMR8_ALL.INTChemical Shifts : 8.0 (1861611 pts)

280260240220200180160140120100806040200-20-40Chemical Shifts : Value (ppm)

-40

-20

0

20

40

60

80

100

120

140

160

180

200

220

240

260

280

Che

mic

al S

hifts

: 8.

0

Version 8.00Version 8.00 Version 10.05Version 10.05

18

Prediction Accuracy

Standard Error of Prediction Formula:

n-1

(δexpi - δcalci)2

n(n-1)

δexpi - δcalci2

-Standard Error of Prediction

=

19

Prediction Accuracy

CNMR Predictor Standard ErrorVersion 8 - 3.11 ppmVersion 9 - 2.32 ppmVersion 10.00 - 2.26 ppmVersion 10.05 – 1.84 ppm

A 21% increase in accuracy over version 9!A 41% increase in accuracy over version 8!

20

Prediction Accuracy

Comparison of HOSE and Neural Network>187,000 chemical shifts used in testNN algorithms- 12% accuracy increase over version 9 HOSE CodeVersion 10 HOSE code- 16% accuracy increase over version 9 HOSE code

HOSE Code is better for now

21

The Future of Neural Nets

What is planned for NMR Predictors?How do Neural Networks fit into these plans?

22

The Future of Neural Nets

Version 11 will further integrate the Neural Network Algorithm

An intelligent hybrid approachMuch like the use of incremental scheme today

Stay tuned for more validation results1H NMR validation study

23

Acknowledgements

Kirill BlinovMikhail KvashaMarina Solnetseva and the database teamRyan Sasaki