Where charge comes from
A peptide has two ionisable termini and one ionisable side chain for every aspartate, glutamate, cysteine, tyrosine, histidine, lysine and arginine in the sequence. Each of those groups is an acid or a base with its own pKa, and each is partly ionised at any given pH.
The net charge is the sum of those fractional contributions. It is not an integer, and treating it as one is the main reason hand estimates disagree with the calculation.
Reading the titration curve
The curve falls monotonically from strongly positive at low pH, where every basic group is protonated and every acid is neutral, to strongly negative at high pH, where the reverse holds. The steepest parts of the curve are where pKa values cluster.
The point where it crosses zero is the isoelectric point. A curve that crosses steeply belongs to a peptide whose charge is sensitive to small pH changes; one that crosses shallowly belongs to a peptide that is relatively forgiving.
What charge predicts
Solubility, first. Charged molecules repel each other, and that repulsion is what keeps them in solution. At the pI the repulsion vanishes and aggregation becomes much more likely, which is why the pI is a pH to avoid rather than a pH to target.
Chromatographic behaviour, second. Cation exchange resins retain positively charged peptides, anion exchange resins retain negatively charged ones, and the pH at which you load determines which. Third, surface interaction: a strongly cationic peptide binds to glass and to negatively charged plastics, which shows up as unexplained loss during handling.
- •Keep the working pH at least one unit from the pI for solubility.
- •Load a cation exchange column below the pI, an anion exchange column above it.
- •Expect adsorptive losses from strongly cationic peptides in standard labware.
- •Histidine is the one residue whose ionisation changes appreciably across the physiological range, since its pKa sits near 6.
The limits of the model
The calculation treats every ionisable group as independent, which they are not. A lysine next to another lysine finds it harder to hold its proton because of the neighbouring positive charge, so its effective pKa falls. The same effect operates in reverse between opposite charges.
The error is small in long, sparsely charged sequences and largest in short, densely charged ones, which unfortunately describes many research peptides. Treat the output as a good estimate of the shape of the curve and an approximate location for the pI, not as a measurement.
How net charge and pI are calculated
Henderson-Hasselbalch applied once per ionisable group and summed, then a bisection search for the pH at which the sum is zero.
basic group contributes + 1 / (1 + 10^(pH - pKa)) acidic group contributes - 1 / (1 + 10^(pKa - pH)) net charge = SUM over all ionisable groups pI = pH where net charge = 0
- Enumerate the ionisable groups. The N-terminal amine, the C-terminal acid, and one group per Asp, Glu, Cys, Tyr, His, Lys and Arg. The table lists each with its pKa and its contribution at the current pH.
- Compute a fractional charge per group. A group whose pKa equals the pH is half ionised and contributes 0.5. Two units away it is 99 percent ionised or 99 percent neutral.
- Sum with sign. Basic groups add, acidic groups subtract. The result is almost never a whole number, which is correct: charge is a population average across many molecules.
- Sample the curve. The same sum is evaluated at 141 points from pH 0 to 14 to draw the titration curve, so the shape as well as the value is visible.
- Bisect for the isoelectric point. Because net charge falls monotonically with pH, bisection converges reliably. A hundred rounds narrow the interval far below the precision of the pKa values themselves.
What this method cannot tell you
- •Ionisable groups are treated as independent. Neighbouring charges shift each other's pKa, and the error is largest in short, densely charged peptides.
- •One published pKa set is used throughout. Other sets differ by up to a pH unit at the termini, which moves the calculated pI by a few tenths.
- •Ionic strength is not modelled. High salt screens charges and shifts effective pKa values.
- •Terminal modifications are invisible. An amidated C-terminus and an acetylated N-terminus each remove an ionisable group, which changes the curve substantially in a short peptide.
Where the numbers come from
Peptide charge at pH: frequently asked questions
Sum a fractional charge for every ionisable group, where each fraction comes from the Henderson-Hasselbalch relationship between the pH and that group's pKa.
The groups are the two termini plus one for every Asp, Glu, Cys, Tyr, His, Lys and Arg in the sequence.
Because it is a population average. At any pH some fraction of the molecules have a given group protonated and the rest do not, and the reported charge is the mean over that population.
A charge of plus 2.3 means the average molecule carries 2.3 elementary charges, not that any individual molecule does.
The pH at which an ionisable group is exactly half ionised. Below it the group is mostly protonated, above it mostly deprotonated.
The transition is not sharp: it takes about four pH units to go from 1 percent to 99 percent ionised, centred on the pKa.
At pH 7.4, aspartate and glutamate are essentially fully negative, lysine and arginine essentially fully positive, and histidine is partly positive.
- •Aspartate, pKa 3.65: negative
- •Glutamate, pKa 4.25: negative
- •Histidine, pKa 6.0: about 4 percent positive at pH 7.4
- •Lysine, pKa 10.53: positive
- •Arginine, pKa 12.48: positive
Because its pKa near 6 is the only side chain pKa inside the physiological range, so its ionisation actually changes as pH moves between 6 and 8.
Every other side chain is fully ionised or fully neutral across that window. Histidine is why a peptide's charge can be genuinely pH-sensitive near neutrality.
Net charge as a function of pH, from 0 to 14. It falls monotonically, steeply where pKa values cluster and shallowly where they do not.
The point at which it crosses zero is the isoelectric point, and the steepness at that crossing tells you how sensitive the peptide's charge is to a small pH change.
Charged molecules repel each other, and that repulsion is what keeps them dispersed. As the net charge approaches zero the repulsion disappears and aggregation becomes far more likely.
This is why the pI is a pH to avoid. Keeping the buffer at least one pH unit away from it usually restores enough net charge to keep the peptide in solution.
One that gives the peptide the charge your resin retains. Cation exchange needs a positively charged peptide, so load below the pI; anion exchange needs the opposite, so load above it.
The further from the pI, the stronger the binding, and the more salt or the larger a pH shift is needed to elute.
It changes the effective pKa values by screening charges, which shifts the curve. The calculator does not model ionic strength.
The effect is modest at physiological salt concentrations and larger in very high or very low salt. It matters most for the position of the pI in a densely charged peptide.
It removes the negatively charged C-terminal carboxylate entirely, so the peptide is one unit more positive across most of the pH range and the pI moves up.
In a short peptide, where the termini are a large share of the total ionisable groups, this can move the pI by more than a full pH unit. The calculator assumes a free acid.
Because they are measured in different model compounds and different conditions, and the terminal values in particular depend heavily on the neighbouring residue.
The spread is up to a pH unit at the termini. Volta uses one set across all its tools so its own figures agree with each other; comparing against another site's pI compares the pKa sets as much as the peptides.
Every peptide with a free N-terminus and a free C-terminus has one, because it is positive at low pH and negative at high pH and must cross zero in between.
A peptide with both termini blocked and no ionisable side chains would have no net charge at any pH, and therefore no meaningful pI.
It constrains it. A highly charged peptide does not cross a lipid bilayer passively, which is one reason peptides are generally poorly absorbed.
Cationic cell-penetrating peptides get in by other routes, chiefly by binding to negatively charged surface molecules and triggering uptake, which depends on more than net charge alone.
Because only the twenty standard residues have defined pKa values here. B, J, O, U, X and Z are discarded rather than guessed.
They used to be counted in the residue total while contributing nothing to the charge, which made the residue count disagree with the number of groups in the table.
Good enough to choose a buffer or a column, not good enough to quote as a measurement. The independent-groups assumption is the main source of error.
For a long sequence with well-separated charges it is close. For a short, densely charged peptide the real curve is flatter than the model predicts.
Anything comfortably away from zero. A net charge of plus or minus two or more at your working pH usually provides enough repulsion to keep the peptide dispersed.
The sign matters less than the magnitude, though it does determine which surfaces the peptide will stick to.
Related Products
Related Research News
Pharmacists Issue Timing Guidance for GLP-1 Users on Birth Control
Pharmacists are advising people who use GLP-1 medications on how to time their doses to avoid interactions with birth control. The guidance, reported by Yahoo Health on September 12, 2026, focuses on when to take each product so contraceptive effectiveness is not compromised.
BPC-157 for Dogs: Benefits, Research and the Two Canine Studies
Go through the BPC-157 literature looking for studies that involved dogs and you find two: one toxicology, one pharmacokinetics. Neither asks whether it heals anything. What the canine evidence actually contains, why the published dosing charts disagree by more than 20-fold, and what the 2026 reviews concluded.
Tesamorelin vs Ipamorelin: Human Evidence, Doses, Safety
Tesamorelin holds FDA approval for HIV-associated lipodystrophy; ipamorelin has none. This page maps the published human dosing data for each and marks the gaps.