rapid measurement of corn amino acids using near infrared ... · level levels of selected amino...

Value Added Agriculture ProgramValue Added Agriculture Program

Rapid Measurement of Corn Amino Acids using Near Infrared Whole Grain Analyzers

Connie L Hardy Glen R RippkeConnie L. Hardy, Glen R. RippkeCharles R. Hurburgh, Walter A. Goldstein

AgMRC Special Project 2007-2008

ISU Extension Iowa Grain Quality InitiativeMichael Fields Agricultural Institute

ISU Extension Value Added Agriculture Program

www.iavaap.org


ObjectivejDevelop rapid method to measure important

amino acids in corn for poultry and p ylivestock rations using near-infrared spectroscopy (NIRS) whole grain analyzersanalyzers

Why?Why?• Current lab methods are costly and time consuming• Corn breeding programs and feed ration balancing will

be facilitated with rapid measurement systems

www.iavaap.org

be facilitated with rapid measurement systems


NIRS measures light energyNIRS measures light energy absorption

www.iavaap.org


How NIRS analyzers work…

3

3.05

3.1

3.15

0 10 20 30 40 50 60 70 80 90 1002.75

2.8

2.85

2.9

2.95

kk xbxbxbbY ++++= ...ˆ22110

Prediction model

Protein content: 14 g.kg-1

www.iavaap.org


Can’t we do this already?• Calibration attempts to measure selected

amino acids in grain usually predict total g y pprotein and only calculate a typical amino acid level based on the protein plevel.

• Accurate measurement of amino acidsAccurate measurement of amino acids may rely primarily on the characteristics of the samples used for calibration

www.iavaap.org

the samples used for calibration.


Good samples are the key t !

The corn samples available from breeding programs di t d b Mi h l Fi ld A i lt l I tit t

to success!coordinated by Michael Fields Agricultural Institute represent unique ratios of selected amino acids(methionine lysine cysteine)(methionine, lysine, cysteine) to total protein; thus the analyzer can “learn” this new relationship pand provide a more accurate measurement.

www.iavaap.org


What did we do?• We used breeder samples with relatively high

level levels of selected amino acids fromlevel levels of selected amino acids from harvest years 2006 (106 samples) and 2007 (406 samples)( p )– Lysine 0.26-0.53– Methionine 0.14-0.39Methionine 0.14 0.39– Cysteine 0.14-0.37

www.iavaap.org


What else did we do?• Scanned samples in two whole seed

transmission analyzers Bruinstransmission analyzers - Bruins OmegAnalyzer G and Foss Infratec™ 1241 Grain Analyzer1241 Grain Analyzer

• Used CAMO Unscrambler v.9.7 software

www.iavaap.org


Methionine

R2 = 0.75

Standard Error of CrossStandard Error of Cross Validation (SECV) = 0.029

Range: 0.14 – 0.39

www.iavaap.org


LysineLysine

R2 = 0.84

Standard Error of Cross Validation (SECV) = 0.021

Range: 0 26 – 0 53Range: 0.26 – 0.53

www.iavaap.org


Cysteine

R2 = 0.78

Standard Error of Cross Validation (SECV) = 0.016

Range: 0.14 – 0.37

www.iavaap.org


Spectral correlations with i id t t l t iamino acids vs. total protein

R2 Determination withRange Omega Infratec Total(%db) Spectra Spectra Protein

MET 0.14-0.39 0.746 0.730 0.542LYS 0.26-0.53 0.837 0.842 0.390CYS 0.14-0.37 0.783 0.787 0.797

Kovalenko et al. 2006 proposed using R2 or RPD (Relative Predictive Determinant) to test whether values are calculated values or true predicted values

www.iavaap.org

values are calculated values or true predicted values.


What do the results mean?• Methionine and lysine can be measured by NIRS with

ffi i t f ti isufficient accuracy for genetic screening.• Measurement of cysteine reflected protein

measurement therefore we cannot conclude that we aremeasurement, therefore we cannot conclude that we are able to measure cysteine independently.

• Additional carefully selected calibration samples from future harvests will likely enhance accuracy and serve as a validation set.

www.iavaap.org


Thanks! to our sponsor and llcolleagues

www.iavaap.org

rapid measurement of corn amino acids using near infrared ... · level levels of selected amino...

Documents