Ultraviolet monitoring of chromatography performance by orthogonal partial least squares

Patent metadata

Field	Value
Publication number	US-11885779-B2
Application number	US-202117229186-A
Country	US
Kind code	B2
Filing date	Apr 13, 2021
Priority date	Apr 14, 2020
Publication date	Jan 30, 2024
Grant date	Jan 30, 2024

Abstract

Disclosed are methods for monitoring column chromatography performance. The methods can include acquiring one or more chromatogram ultraviolet (UV) traces generated by a chromatography system during sample purification and/or separation; and analyzing the one or more acquired chromatogram UV traces with an orthogonal partial least squares (OPLS) model, thereby allowing detection of column deterioration prior to column failure and quantitative analysis of UV signal in the one or more chromatogram UV traces.

First claim

What is claimed is:

1. A method of detecting chromatography column deterioration prior to column failure in a chromatography system, comprising:
(i) creating an orthogonal partial least square (OPLS) model, wherein creating the OPLS model comprises:
(a) selecting a process or unit operation for which the OPLS model is being created,
(b) collecting raw data for ultraviolet (UV) chromatogram traces of one or more column cycles/lots/runs for the selected process or unit operation,
(c) normalizing and aligning the collected raw data,
(d) generating artificially created curves from normalized raw data,
(e) classifying and formatting data for importation into a multivariate tool,
(f) importing classified and formatted data into the multivariate tool to generate a training set of chromatogram traces, wherein the artificially created curves are classified as “undesirable” chromatogram traces in the training set,
(g) generating the OPLS model from the multivariate tool using the training set,
(h) optimizing and validating the generated OPLS model; and

(ii) acquiring one or more chromatogram ultraviolet (UV) traces generated by a chromatography system during the selected process or unit operation; and
(iii) applying the OPLS model to the one or more chromatogram UV traces, thereby allowing detection of column deterioration prior to column failure.

2. The method of claim 1 , further comprising testing the optimized OPLS model.

3. The method of claim 1 , wherein the process or unit operation is (a) a chromatography unit operation for a single molecule, or (b) a protein affinity chromatography step for the single molecule.

4. The method of claim 1 , wherein:
(a) collecting raw data for a UV trace of one or more column cycles/lots/runs for the selected process or unit operation comprises collecting UV absorbance values at corresponding column volumes for the one or more cycles/lots/runs; and/or
(b) normalizing and aligning the collected raw data comprises normalizing and aligning UV absorbance values and aligning column volumes.

5. The method of claim 4 , wherein normalizing UV absorbance values comprises removing variation in magnitude differences in UV raw signal.

6. The method of claim 1 , further comprising providing a sample to the chromatography system prior to acquiring the one or more chromatogram ultraviolet (UV) traces generated by a chromatography system during sample purification and/or separation.

7. The method of claim 1 , wherein the chromatography system is a liquid chromatography system.

8. The method of claim 6 , wherein the sample comprises:
(a) a protein;
(b) an antibody, a fusion protein, recombinant protein, or a combination thereof;
(c) a monoclonal antibody; or
(d) a monoclonal antibody of isotype IgG1, IgG2, IgG3, IgG4, or mixed isotype.

9. A method of creating an orthogonal partial least square (OPLS) model for detection of chromatography column deterioration, comprising:
(a) selecting a process or unit operation for which the OPLS model is being created;
(b) collecting raw data for one or more ultraviolet (UV) chromatogram traces of one or more column cycles/lots/runs for the selected process or unit operation;
(c) normalizing and aligning the collected raw data;
(d) generating artificially created curves from normalized raw data;
(e) classifying and formatting data for importation into a multivariate tool;
(f) importing classified and formatted data into the multivariate tool to generate a training set, wherein the artificially created curves are classified as “undesirable” chromatogram traces for the training set; and
(g) generating the OPLS model from the multivariate tool using the training set.

10. The method of claim 9 , wherein the process or unit operation is (a) a chromatography unit operation for a single molecule, or (b) a protein affinity chromatography step for the single molecule.

11. The method of claim 9 , wherein:
(a) collecting raw data for a UV chromatogram trace of one or more column cycles/lots/runs for the selected process or unit operation comprises collecting UV absorbance values at corresponding column volumes for the one or more cycles/lots/runs; and/or
(b) normalizing and aligning the collected raw data comprises normalizing and aligning UV absorbance values and aligning column volumes.

12. The method of claim 11 , wherein normalizing UV absorbance values comprises removing variation in magnitude differences in UV raw signal.

13. The method of claim 9 , wherein the one or more UV chromatogram traces is generated by a chromatography system during sample purification and/or separation.

14. A non-transitory computer-readable storage medium with an executable program stored thereon for detecting chromatography column deterioration in a chromatography system, wherein the program instructs a microprocessor to perform the steps of:
(i) creating an orthogonal partial least square (OPLS) model, wherein creating the OPLS model comprises:
(a) selecting a process or unit operation for which the OPLS model is being created,
(b) collecting raw data for one or more ultraviolet (UV) chromatogram traces of one or more column cycles/lots/runs for the selected process or unit operation,
(c) normalizing and aligning the collected raw data,
(d) generating artificially created curves from normalized raw data,
(e) classifying and formatting data for importation into a multivariate tool,
(f) importing classified and formatted data into the multivariate tool to generate a training set of chromatogram traces; wherein the artificially created curves are classified as “undesirable” chromatogram traces in the training set;
(g) generating the OPLS model from the multivariate tool using the training set; and
(h) optimizing and validating the generated OPLS model, and

(ii) acquiring one or more chromatogram ultraviolet (UV) traces generated by a chromatography system during sample purification and/or separation; and
(iii) analyzing the one or more acquired chromatogram UV traces with the OPLS model, thereby allowing detection of column deterioration prior to column failure and quantitative analysis of UV signal in the one or more chromatogram UV traces.

15. The non-transitory computer-readable storage medium of claim 14 , wherein the process or unit operation is (a) a chromatography unit operation for a single molecule, or (b) a protein affinity chromatography step for the single molecule.

16. The non-transitory computer-readable storage medium of claim 14 , wherein:
(a) collecting raw data for a UV trace of one or more column cycles/lots/runs for the selected process or unit operation comprises collecting UV absorbance values at corresponding column volumes for the one or more cycles/lots/runs; and/or
(b) normalizing and aligning the collected raw data comprises normalizing and aligning UV absorbance values and aligning column volumes.

17. The non-transitory computer-readable storage medium of claim 16 , wherein
normalizing UV absorbance values comprises removing variation in magnitude differences in UV raw signal.

18. The non-transitory computer-readable storage medium of claim 14 , further comprising providing a sample to the chromatography system prior to acquiring the one or more chromatogram ultraviolet (UV) traces generated by a chromatography system during sample purification and/or separation.

Assignees

• Regeneron Pharma

Inventors

• Shierly Eric
• Brennan Kevin
• Mao Nathan

Classifications

• G01N30/74 Primary

  Optical detectors {(measurement of intensity, velocity, spectral content, polarisation, or phase of infrared, visible or ultraviolet light G01J)} · CPC title

• G01N1/34

  Purifying; Cleaning {(processes or apparatus for extracting or separating nucleic acids from biological samples C12N15/1003)} · CPC title

• G01N21/33

  using ultraviolet light (G01N21/39 takes precedence) · CPC title

• G01N30/04

  Preparation or injection of sample to be analysed · CPC title

• G01N30/88

  Integrated analysis systems specially adapted therefor, not covered by a single one of the groups G01N30/04 - G01N30/86 · CPC title