Preparing inputs

A tool's input requirements describe both the file format and the data it can analyze. A valid FASTA file, for example, can still contain the wrong sequence type or too many records for a particular tool.

Start with the input description on the chosen tool page. Check its accepted formats, required inputs, and limits before preparing your file.

Sequences

FASTA contains a header beginning with > followed by the sequence on the next line. This synthetic protein example contains the 20 standard amino acids:

>demo_protein
ACDEFGHIKLMNPQRSTVWY

Use short, distinct record names so outputs can be traced back to their inputs. For multiple records, repeat the header-and-sequence pair. Add multiple records only when the tool supports them; some tools require a single sequence, while others treat several sequences as a joint input.

• Protein, DNA, or RNA: Select the sequence type the tool expects. A filename does not establish the molecule type.
• Ambiguous residues: Check whether the tool accepts ambiguity codes, nonstandard residues, or stop symbols. Do not silently replace them to make validation pass.
• Copied text: Paste sequence content without line numbers, page headings, or formatting from a document editor.
• Length and count: Check limits for each record and the whole submission. Passing a format check does not bypass a tool's scientific limits.

The start guide uses the example above for a complete first analysis.

Alignments

An alignment records corresponding positions across sequences, including gaps. An ordinary file containing several unaligned sequences is not an alignment merely because it uses FASTA.

When a tool requires an alignment, supply one produced by an alignment method and preserve its gap characters and format. When a tool requires an ungapped sequence, check its instructions before removing gaps. Some specialized formats carry additional information that is lost by treating them as plain FASTA.

The MSA Viewer can inspect an existing alignment. It does not align sequences for you.

Structures

PDB and mmCIF files contain molecular coordinates and identifiers. Download the actual structure file rather than a webpage displaying the structure, and keep the original file before making changes.

Check the parts that matter for your analysis:

• Chains and models: Confirm which chains and coordinate models are present and which the tool will use.
• Residues and atoms: Check for missing residues, incomplete atoms, or unsupported residue types.
• Ligands, ions, and water: Decide what should remain based on the analysis. Removing all nonprotein atoms can remove biologically relevant material.
• Hydrogens and preparation: Follow the destination tool's requirements. Preparation settings can change the structure being analyzed.

Use a viewer to inspect the file and a preparation tool when you need to modify it. Renaming .cif to .pdb does not convert its contents.

Small molecules

Small-molecule tools may accept a text representation such as SMILES or a structure file such as SDF. Supply the representation requested by the input field. A database identifier is usable directly only when the tool offers that input method.

Check stereochemistry, charge, and the number of molecules before submission. A representation of chemical connectivity does not necessarily contain the three-dimensional coordinates needed by a downstream calculation.

Molecule Drawer can create or edit a molecule and export supported representations. Use a preparation or conversion tool when the destination requires additional processing.

Files and repeated analyses

Files accepts supported, nonempty files up to 50 MB each, subject to your storage allowance. Individual tools can impose additional or lower limits. Uploading a file to the library does not mean every tool accepts it.

When using Send to tool, review the destination input assignments. Format compatibility helps find a destination; you still need to check the scientific content and settings.

For repeated analyses, use Batches and inspect the preview before submission. Do not assume that one file always creates one job: the chosen tool determines how inputs are grouped.

Resolve an input error

Use the reported field or record to locate the problem. Check the contents in a plain-text editor, confirm the selected format, and correct the smallest relevant issue. Keep a copy of the original data and record meaningful changes.

If the input passes validation but the run fails, use the job's error details to troubleshoot the analysis. Format validation cannot establish that the chosen method is appropriate for the data.