Aliphatic Index

How to calculate a protein's aliphatic index

Paste a protein sequence or upload a sequence file, then select Generate to calculate its aliphatic index. ProteinIQ reports the sequence length and a composition-based index weighted by the side-chain volumes of alanine, valine, isoleucine, and leucine. Submit several sequences to compare proteins or variants in one results table.

For a single sequence, paste the mature human insulin B chain:

FVNQHLCGSHLVEALYLVCGERGFFYTPKT

The result is labeled Sequence, with 30 amino acids and an aliphatic index of 84.33.

To compare both mature insulin chains, give each sequence its own FASTA header:

>insulin-B human insulin B chain
FVNQHLCGSHLVEALYLVCGERGFFYTPKT
>insulin-A human insulin A chain
GIVEQCCTSICSLYQLENYCN

The calculation returns these values (the table initially sorts by Protein ID):

The value includes every residue in the sequence length. For example, a residue that is not A, V, I, or L contributes zero to the numerator, but still contributes to the denominator.

Input

Sequence lines are joined, whitespace is ignored, and lowercase amino acid codes are converted to uppercase. Non-standard residues, alignment gaps, stop markers, digits, and other unsupported sequence characters are rejected rather than removed. An empty FASTA record is also rejected. Put each FASTA sequence on the line after its header; descriptions stay in the header and never contribute to the calculation. For example, >insulin-B human insulin B chain appears as insulin-B in the results table.

CSV files

CSV files need a header row with id and sequence columns. Common column names such as name and seq are also recognized. Quoted cells are supported. Save the following as a .csv file and upload it through Files to obtain the same insulin results shown above:

id,sequence
insulin-B,FVNQHLCGSHLVEALYLVCGERGFFYTPKT
insulin-A,GIVEQCCTSICSLYQLENYCN

Each data row produces one result. A missing sequence, incomplete row, or unsupported sequence character stops the calculation rather than silently dropping that row. Keep identifiers unique and free of spaces or pipes when you need the complete identifier in the result.

PDB structures

PDB files produce one result for each protein chain in the first model, using residues present in the coordinates. Unresolved residues and non-protein chains are not included, so the result can differ from the full deposited sequence. Recognized modified residues use the existing PDB-to-FASTA mappings, such as MSE to methionine. Unknown protein residues are rejected. Use PDB to FASTA first if you need to choose a different sequence source or inspect the extracted sequences.

Input limits

Sequence input is limited to 5 records per run for guests and 100 for Free accounts. Paid plans have no configured sequence-count cap. The 50 MB upload limit still applies.

Settings

This calculator has no user-configurable settings. It always uses the Ikai coefficients of 2.9 for valine and 3.9 for isoleucine plus leucine.

Results

Each sequence produces one row. The table initially sorts by Protein ID; select a column heading to change the sort order. Invalid input stops the run, so a failed batch does not return a partial results table.

What is the aliphatic index?

The aliphatic index estimates the relative volume occupied by the aliphatic side chains in a protein. It was introduced by Ikai as a sequence-derived property for comparing globular proteins, and is often used as one indicator when investigating thermal stability.

The calculation uses the Ikai formula documented by ExPASy:

Here, each $X$ is the mole percentage of that residue in the complete sequence. The weights mean that a given percentage of valine contributes 2.9 times as much as alanine, while isoleucine and leucine each contribute 3.9 times as much. A sequence made entirely of valine therefore has an index of 290, whereas one made entirely of isoleucine or leucine has an index of 390.

How should I interpret an aliphatic index?

Use the result as a comparative property, not a temperature prediction. A higher value means the sequence is more enriched in the four weighted aliphatic residues. That can be informative when comparing matched homologs, constructs, or single-site variants, especially when sequence length and protein class are similar.

It does not account for tertiary structure, oligomerization, disulfide bonds, cofactors, solvent conditions, pH, or post-translational modifications. A high index alone does not establish that a protein will express well, remain soluble, or tolerate a particular temperature. For a mutation series, compare the index with measured stability and other sequence properties rather than treating a cutoff as a pass or fail rule.

Which protein property tool should I use?

FAQ

What amino acids are used in the aliphatic index?

Alanine, valine, isoleucine, and leucine. Their mole percentages are weighted as 1.0, 2.9, 3.9, and 3.9, respectively; all other residues affect the sequence length but add no weighted contribution.

Is a higher aliphatic index always better?

No. It only indicates greater representation of the weighted aliphatic residues. Whether that is desirable depends on the protein, target conditions, and trade-offs with folding, solubility, activity, and expression.

Can I calculate the aliphatic index for multiple proteins at once?

Yes. Paste multiple FASTA records or upload a supported file. Each sequence is calculated separately, and the results table starts sorted by Protein ID. All sequences must pass validation for the run to return results.

Can the aliphatic index be greater than 100?

Yes. It is a weighted index, not the percentage of aliphatic residues. The formula ranges from 0 to 390: an all-valine sequence scores 290, while an all-isoleucine or all-leucine sequence scores 390.

Why is my protein sequence rejected?

The calculator accepts only the 20 standard amino acid letters. Ambiguous residues such as X, B, and Z, alignment gaps, stop markers such as *, and residue numbering must be resolved before calculation. It does not delete these characters or guess their identities.

Why does my FASTA header look different in the output?

The results table uses the first token in the header, stopping at whitespace or |. Name a record >variant-7 description when variant-7 is the identifier you want to compare and export.