Determination of chromatography conditions

An apparatus for determining isocratic chromatography conditions for separating a biomolecule from a liquid sample, the apparatus comprising,

• a computer configured to communicate with a software to execute a set of instructions to:
• a) select a number of experiments (/V experiments) using design of experiments (DoE);
• b) output to a controller for carrying out a number of 44 chromatography runs (M chromatography runs) for said N experiments on a chromatography column with in-line conditioning of at least two buffer gradients of two selected quality measures wherein paths and curvatures of the gradients are designed so that the points in the DoE are close to the paths of the gradients to obtain N elution fractions during the M chromatography runs, one for each of the DoE points passed by;
• c) analyze a purity and yield of the N elution fractions to obtain responses data;
• d) use the responses data to train a DoE regression model;
• and
• e) use the DoE regression model to predict isocratic chromatography conditions for said biomolecule based on the results from b);
• wherein the two selected quality measures are selected from the group consisting of type of chromatography resin;
• loading volume, wash volume, elution volume, binding buffer;
• elution buffer;
• surface determinants of the biomolecule and impurities present in the liquid sample and are orthogonal to each other, and
• wherein the number of chromatography runs (M) is less than the number of experiments of DoE (N).

A method for determination of chromatography conditions for separating a biomolecule from a liquid sample, the method comprising,

• a1) preliminarily selecting a number of DOE experiments (N DoE experiments) using multivariate analysis,
• a2) adjusting experimental parameters of said experiments to fall on at least two gradients of two selected quality measures,
• b) performing a number of chromatography runs (M chromatography runs) for said N experiments on a chromatography column with in-line conditioning of at least two gradients of two selected quality measures wherein paths and curvatures of the gradients are designed so that the points in the DoE are close to the paths of the gradients to obtain N elution fractions during the M chromatography runs, one for each of the DoE points passed by;
• c) analyzing a purity and yield of the N elution fractions to obtain responses data;
• d) using the responses data to train a DoE regression model;
• and,
• e) using the DoE regression model to predict isocratic chromatography conditions for said biomolecule;
• wherein the two selected quality measures selected from the group consisting of type of chromatography resin;
• loading volume, wash volume, elution volume, binding buffer;
• elution buffer;
• surface determinants of the biomolecule and impurities present in the liquid sample and are orthogonal to each other,
• wherein at least two of said at least two gradients are curved gradients, wherein the number of chromatography runs (M) is less than the number of experiments of DoE (N),
• wherein the in-line conditioning comprises conditioning of a liquid sample to a predetermined pH and/or conductivity, and
• wherein the biomolecule is a monoclonal antibody, or a fragment or fusion.