Protein molecular weight calculator

How to calculate protein and peptide molecular weight

Paste a protein or peptide sequence, a FASTA file, or a CSV table, and ProteinIQ returns the molecular weight of each chain in daltons (Da) and kilodaltons (kDa). The calculation sums residue masses and adds one water, so a 5-residue peptide and a 1,000-residue protein go through the same formula in the browser.

For a small molecule, enter its SMILES in Molecular descriptors instead, which reports the formula, molecular weight and exact mass.

Human ubiquitin, pasted as FASTA with the default settings:

>Human_ubiquitin
MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPDQQRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGG

The same calculator handles a peptide. Leu-enkephalin (YGGFL) comes out at 555.631 Da average, or 555.269 Da monoisotopic, which is the number to look for in a mass spectrum. Synthetic peptides with an acetylated N-terminus or an amidated C-terminus have their own settings, covered in peptide molecular weight.

Accepted input

Ambiguous codes use averaged masses (B between Asn and Asp, Z between Gln and Glu) and X uses 110 Da, so a sequence that contains them gives an approximate result. Any other character, including * stop symbols, is rejected with the offending letters named in the error.

A CSV table looks like this:

id,sequence
p1,MAG
p2,ACD

Settings

Leave disulfide bonds at 0 for a reduced protein, which is what a reducing SDS-PAGE gel or a reduced LC-MS run measures. The terminus settings apply to every sequence in the input, so split a batch that mixes capped and free peptides into two runs.

Results

Each sequence produces one row:

One dalton per molecule equals one gram per mole, so 8564.845 Da also means 8564.845 g/mol when you weigh out ubiquitin for a molar solution.

How protein and peptide molecular weight is calculated

Every peptide bond releases one water molecule. Join the five free amino acids of Leu-enkephalin and four waters leave, so the peptide weighs 72.061 Da less than its building blocks:

Counting $n-1$ waters for $n$ amino acids is easy to get wrong, so the calculator uses residue masses instead. A residue mass is the amino acid minus one water, which leaves exactly one water to add back for the free N- and C-termini:

Here $m_i$ is the residue mass of each amino acid, $m_{H_2O}$ is 18.01528 Da (average) or 18.01056 Da (monoisotopic), $k$ is the number of disulfide bonds, and $\Delta_N$ and $\Delta_C$ are the terminal cap shifts, zero for free termini. Both routes give 555.631 Da for Leu-enkephalin.

The average residue masses used by default:

Average atomic masses follow the IUPAC standard atomic weights, and the residue convention matches the ExPASy protein analysis tools.

Average vs monoisotopic mass

Carbon is about 1.1% carbon-13, so a protein with hundreds of carbons almost never consists of the lightest isotopes alone. Average mass weights every isotope by its natural abundance and describes the bulk material you weigh or load on a gel. Monoisotopic mass counts only ¹²C, ¹H, ¹⁴N, ¹⁶O and ³²S, which is the first peak of an isotope envelope.

The gap grows with size. For a peptide the 0.36 Da difference decides whether a high-resolution MS peak matches, so use monoisotopic mass for peptides and isotope-resolved spectra. For an intact protein the monoisotopic peak is small or invisible, and deconvoluted intact-mass results are usually compared with the average mass. Gel, concentration and buffer calculations always use average mass.

Peptide molecular weight

Oxytocin is nine residues long, and two parts of its structure change the mass its sequence alone would give. Its two cysteines close a ring through a disulfide bond, and its C-terminal glycine ends in an amide rather than a free acid. Paste CYIQNPCLG with the default settings and the calculator returns 1,010.191 Da. Set Disulfide bonds to 1 and C-terminus to Amidated (-NH2), as the Oxytocin peptide example does, and it returns 1,007.190 Da.

The values below were calculated here and checked against PubChem, which lists oxytocin at 1,007.2 g/mol with an exact mass of 1,006.4365 Da:

A peptide certificate of analysis normally quotes the average molecular weight, which is the number for weighing out a solution. Peptide powder also carries counterions such as trifluoroacetate and some water, so a milligram of powder holds less than a milligram of peptide; vendors report the difference as net peptide content.

In a mass spectrum of a peptide this size the monoisotopic peak is the tallest, and a singly protonated ion appears at the monoisotopic mass plus 1.007 Da. To list the ion m/z values of every fragment in a protein digest instead, use the peptide mass calculator, which cuts a protein with a chosen protease and reports each peptide as [M+H]⁺, [M+2H]²⁺ or another ion.

Amino acids to kDa: the 110 Da rule

A residue weighs about 110 Da on average, so multiplying the number of amino acids by 0.11 gives a quick estimate in kDa. In reverse, a protein has about 9 residues per kDa, which puts a 50 kDa band at roughly 455 amino acids.

The shortcut is good to a few percent, which is enough to pick a gel percentage or a size-exclusion column. It drifts when a sequence is rich in heavy residues such as Trp, Arg and Tyr, or light ones such as Gly and Ala, so use the sequence whenever you have it. The average protein size guide lists typical lengths and kDa values across proteomes.

Count everything that is in the expressed chain. A His6 tag adds 822.8 Da of residues and GFP adds about 27 kDa, but only when those residues are part of the pasted sequence, so paste the construct as expressed rather than the native protein.

From kDa to molarity

Molecular weight turns a mass concentration into a molar one. Divide grams per litre by the molecular weight in g/mol: 1 mg/mL of ubiquitin is 1 g/L ÷ 8,564.845 g/mol, or 116.8 µM. For a 50 kDa protein the same 1 mg/mL is 20 µM.

Run it the other way to weigh out a stock: a 1 mM solution of oxytocin needs 1.007 mg per mL. When the concentration comes from absorbance rather than weighing, the extinction coefficient calculator gives the molar value from A280 directly.

Worked example: erythropoietin from gene to glycoprotein

Human erythropoietin (UniProt P01588) shows how far the protein in the tube can be from the sequence in a database. Each step below changes the number the calculator returns, and the last one is a mass no sequence tool can calculate.

Starting from the precursor overstates the secreted polypeptide by about 3 kDa. Paste the mature chain instead, which UniProt lists under "Chain" in the processing features; the Remove initiator Met setting only handles the first methionine, not signal peptides or propeptides.

The polypeptide still accounts for only 60% of the real molecule. Davis and colleagues measured recombinant erythropoietin at 30,400 Da with 39% carbohydrate, from three N-linked glycans and one O-linked glycan. That matches the calculation: 18,236 Da of protein is 60% of 30,400 Da, leaving about 40% sugar.

Post-translational modifications and glycosylation

The calculator returns the mass of the unmodified chain. When you know which modifications are present, add their mass shifts to the result:

Synthetic peptides are where this matters most. A C-terminal amide and an N-terminal acetyl together shift a peptide by +41.027 Da monoisotopic, and a peptide ordered with both caps will not match an uncapped calculation. Those two caps have their own N-terminus and C-terminus settings; add the other shifts in this table by hand.

Glycans are harder because they are heterogeneous. A single N-linked site carries anything from the 892.317 Da (monoisotopic) Man₃GlcNAc₂ core to branched, sialylated structures of 2 to 3 kDa, and each copy of the protein can differ.

Find candidate sites with N-glycosylation site prediction, then treat the glycoprotein mass as a range rather than a single number. Antibodies are a common case, covered in antibody molecular weight.

SDS-PAGE vs calculated molecular weight

A band that sits away from the calculated mass is common and does not by itself mean the protein is wrong. SDS-PAGE estimates mass by comparing mobility with a ladder, which assumes every protein binds the same amount of SDS per gram and unfolds into a similar shape. Several things break that assumption:

• Glycosylation adds mass and binds little SDS. Erythropoietin has 18.2 kDa of polypeptide but weighs 30.4 kDa as a glycoprotein, and its band reflects the glycoprotein, not the chain.
• Membrane proteins and very hydrophobic segments bind more or less SDS than soluble proteins. Rath and colleagues measured shifts of −10% to +30% relative to formula weight for transmembrane hairpins.
• Highly charged or proline-rich sequences, and incompletely reduced disulfides, migrate anomalously.
• Fusion tags, uncleaved signal peptides and proteolytic clipping change the actual chain on the gel.

Gel filtration has its own trap. A size-exclusion column sorts by Stokes radius, not mass, as Erickson reviews. Glycosylated erythropoietin has a Stokes radius of 32 Å, much larger than the 20 Å expected for a compact sphere of its mass, so it elutes like a bigger protein.

When the band and the calculation disagree, check the sequence you calculated against the construct first (tags, signal peptide, initiator Met), then account for known modifications, then confirm by intact mass spectrometry if the difference still matters.

Molecular weight examples

The calculator includes four examples that load the sequence and settings described here.

Beta-lactoglobulin reference sequence

The 162-residue mature bovine chain published with the ExPASy Compute pI/Mw reference output. With Average mass, no initiator Met removal and no disulfide correction, ProteinIQ returns 18,281.208 Da (18.281208 kDa). ExPASy reports 18,281.00 Da; the 0.208 Da difference (about 0.0011%) comes from the precision of each mass table.

Human ubiquitin

The canonical 76-residue sequence with Average mass and no corrections, returning 8,564.845 Da (8.564845 kDa). It is a compact check of raw FASTA input and the default residue table.

Oxidized insulin A-chain

The 21-residue human A-chain with Monoisotopic mass and one intrachain disulfide bond, returning 2,379.984 Da (2.379984 kDa). This is the isolated A-chain only. Intact insulin also contains the B-chain and two interchain disulfides, as described in how many amino acids are in insulin.

Oxytocin peptide

The nine-residue hormone CYIQNPCLG with Average mass, one disulfide bond and an amidated C-terminus, returning 1,007.190 Da. PubChem lists oxytocin at 1,007.2 g/mol. Switch to Monoisotopic and the result, 1,006.436 Da, matches PubChem's exact mass of 1,006.4365 Da.

Which molecular weight tool should I use?

Frequently asked questions

What is the molecular weight of a protein?

It is the sum of the masses of every atom in one molecule, given in daltons (Da) or kilodaltons (kDa). Most proteins fall between 10 and 100 kDa; human ubiquitin is 8.6 kDa and serum albumin about 66 kDa. Paste a sequence above to get the exact value for your protein.

How do I convert amino acids to kDa?

Multiply the number of amino acids by 0.11 for an estimate, so 300 residues is about 33 kDa. For the exact value, paste the sequence into the calculator; the shortcut is typically within a few percent.

How many amino acids is 1 kDa?

About 9, since an average residue weighs 110 Da. A 50 kDa protein therefore has roughly 455 amino acids. The real count depends on composition, and glycosylation or tags can make a smaller chain run at 50 kDa on a gel.

How do I calculate the molecular weight of a peptide?

Add the residue masses and one water (18.015 Da), or equivalently add the free amino acid masses and subtract 18.015 Da for each peptide bond. Paste the peptide here, set N-terminus and C-terminus to match any acetyl or amide caps, enter the disulfide count, and choose Monoisotopic when you compare with a mass spectrum.

How do I convert protein molecular weight in kDa to molarity?

Divide the mass concentration in g/L (the same number as mg/mL) by the molecular weight in g/mol. A 50 kDa protein at 1 mg/mL is 1 ÷ 50,000 mol/L, or 20 µM.

Why is my protein bigger on SDS-PAGE than the calculated molecular weight?

Glycosylation, phosphorylation, uneven SDS binding, fusion tags and uncleaved signal peptides all move a band up. Membrane proteins can also run faster than expected. When the gap is larger than known modifications explain, intact mass spectrometry confirms what the chain really is.

Is a dalton the same as g/mol?

Numerically, yes. A protein of 8,564.845 Da per molecule has a molar mass of 8,564.845 g/mol, which is the value to use for molarity and stoichiometry.

How big is a 100 kDa protein?

For a compact globular protein, Erickson gives a minimum radius of $R_{min} = 0.066 \, M^{1/3}$ nm, with $M$ in daltons. A 100 kDa protein comes out at about 3.1 nm radius, or 6 nm across, and a 50 kDa protein at 2.4 nm radius. Elongated or glycosylated proteins are larger. For a structure file, radius of gyration measures the actual dimensions.

Does the calculator include post-translational modifications?

Only the ones it has settings for: initiator Met removal, disulfide bonds, N-terminal acetylation and C-terminal amidation. Add the shifts from the modification table above for phosphorylation, side-chain acetylation, glycans or other known changes.