It is not possible to predict the solubility of a peptide by simply looking at its structure. But the guanidine of Arg and the e-amino group of Lys may sometimes be used to gauge whether or not a peptide will be soluble, especially if they have short sequences.
Acidic peptides with Glu and Asp, however, tend not to dissolve in water, but will be easily soluble in basic buffers, or when diluted in ammonia. To try and predict the solubility of peptides, here are certain characteristics to consider:
Peptides with less than five amino acids are normally soluble in aqueous solutions. But if the sequence in such peptides, happens to contain hydrophobic residue, then their solubility will be limited to certain solutions, or they may as well be completely insoluble.
Hydrophobic peptides with more than 25% charged amino acids, and less than 25% hydrophobic residues, will easily dissolve in aqueous solutions.
Hydrophobic peptides with 50% or more hydrophobic residues, may dissolve completely or partially in aqueous solutions. Such peptides should be alternatively dissolved in organic solvents, if they do not have the M, W or C residues. If they have these specific amino acids, then they should be dissolved in solutions such as acetic acid, isopropyl alcohol, ethanol, acetonitrile or urea, before they are carefully diluted in any aqueous solution.
Hydrophobic peptides, containing more than 75% hydrophobic amino acids, will generally not dissolve in aqueous solutions. For their dissolution, strong solvents such as formic acid or TFA, have to be used. However, there are chances that precipitation may occur when such peptides are added to the buffered solution, and this may require the use of a high concentration of denaturants or organic solvents to dissolve.
Peptides with a high proportion of D, E, H, K, N, Q, R, S, T or Y may sometimes form intermolecular hydrogen bonds, which can then lead to the formation of gel in the aqueous solution. They should thus, only be dissolved in organic solvents that are compatible with the final experiment.