Choosing the right column, mobile phase for peptide purifica

[pipettemonkey]

I have a couple projects that require me to purify a peptide from insect brains. I have already successfully purified one peptide. I used 6 HPLC steps, exclusively reverse phase (C4, C18, C8, polymer, cyano) to purify the ligand to homogeneity. I frankly feel I got "lucky" with this isolation. I (arbitrarily) picked columns and mobile phase, and had good success moving the active fraction away from impurities.

It seems the choice of column, mobile phase, gradient is so arbitrary (i.e., you don't know how effective a particular system will be until you try it). Is that the case, or is there some logic in deciding which columns, mobile phase to try next?

Also, besides increased signal, what is the advantage of moving from larger to smaller ID columns for later steps?

[Uwe Neue]

There are a bunch of proven procedures around under the umbrella of solid-phase extraction methods that can guide you to a rational approach towards the development of good procedures. I can send you an older article about such procedures.

If you are only interested in a single analyte, the procedures that you use should be optimized to exclude as many things as possible that have significantly different properties than your analyte. For example, you can use the ionic properties of your analyte to separate from anything that does not have ionic properties, meaning that you would use ion exchange as one of the steps. You could use SEC to eliminate stuff that has too small or too large a MW. Both procedures can be rough on/off procedures, but they will cut out a lot of irrelevant garbage, and you are then only concerned about peptides or similar things in a smaller MW range.

Example:
Step 1: Use SEC
Step 2: use on/off cation exchange
Step 3: use on/off anion exchange
Step 4: use and optimize a RP procedure

[pipettemonkey]

There are a bunch of proven procedures around under the umbrella of solid-phase extraction methods that can guide you to a rational approach towards the development of good procedures. I can send you an older article about such procedures.

Thanks. Would you mind sending it to me?

What about switching to smaller ID columns?