Preparing inputs
Choose formats for sequence, structure, and small-molecule files; select chains and regions; and understand what file inspection does and does not check.
Input quality affects computational results. This guide explains supported formats, selection methods, and what file inspection checks.
Input formats
| Category | Format | Contents |
|---|---|---|
| Sequence | FASTA or direct input | Amino-acid or nucleic-acid strings |
| Structure | PDB, mmCIF | 3D atomic coordinates |
| Small molecule | SMILES, SDF | Atoms and bonds, sometimes coordinates |
| Table | CSV | A list for supplying multiple candidates at once |
Supported formats differ by tool. Select a tool to see its required inputs on screen; the tool reference contains the same information.
Preparing data from external sites
- Check format
- Extract archive
- Upload file
- Inspect format
- Select sequence, molecule or chain
Obtain external data directly from its source site. Nadoo Bio does not fetch it on your behalf using identifiers or download URLs. Download access does not necessarily permit commercial computation or processing on external servers; check those terms separately.
- Check the selected tool's accepted formats first. Prepare the data needed for this task rather than an entire database.
- If you downloaded
.gzor.zip, extract it locally before selecting a file. Upload does not automatically decompress or convert formats. - Check the file and format in the project's File upload screen. Extensions below help select a format; file contents are inspected separately.
- After saving, check inspection results and select the needed sequences, molecules, or chains from the inspected data. Recheck your selection when using only part of the original.
| Prepared data | Filename examples | Notes |
|---|---|---|
| Sequence | .fasta, .fa, .faa, .fna | ID lists and search-result tables are not sequence files. Check molecule type and tool-specific length limits. |
| Atomic coordinates | .pdb, .cif, .mmcif | Use text structure files. Renaming a compressed or binary file's extension does not convert it. |
| Molecule list | .sdf, .smi, .smiles | Inspects SDF record boundaries and SMILES notation. A stored list is not necessarily valid input for every tool. |
| Candidate table | .csv | After syntax inspection, check each tool's required columns and value constraints. |
| Multiple sequence alignment | .a3m, CSV for paired alignments | These differ from ordinary FASTA and candidate CSV files. Select the paired-alignment CSV format explicitly when uploading. |
.txt defaults to plain text. If it contains FASTA or SMILES, choose the format matching its contents before upload, or paste into a supported input screen. Renaming a file does not convert its contents.
Storage and inspection have different limits
Storage supports up to 512 MiB, while current format inspection supports 16 MiB per file. One MiB is 1,048,576 bytes. Being able to store a large file does not mean it can pass inspection and be used for computation.
Within 16 MiB, inspection allows up to 10,000 FASTA sequences, 100,000 CSV records including the header, 5,000 SDF molecules, or 500 SMILES entries. Separate limits on individual lengths, atoms, and bonds also apply, so staying below these counts does not guarantee a pass. Execution tools apply their own limits again.
To extract a subset from a large source, use the source site's export options or a local format-aware tool. Do not cut a file mid-sequence or mid-molecule; record the source and selection. Storing plain text, JSON, or binary files does not imply scientific-format inspection support.
Sequence input
FASTA is a text format with repeating >name header lines and sequence lines. One file can contain multiple sequences.
Aligned sequences include markers that match positions across sequences. MSA files contain gap characters such as - and .. These show correspondence, not actual amino acids. Equal sequence lengths alone do not establish that a file is aligned.
Validation reports nonstandard characters and their positions. Correct them and validate again.
Selecting chains and regions in structure files
PDB and mmCIF store atomic coordinates. You usually need to select which chain and region to use. Three distinctions help avoid mistakes.
Residue numbers are not array indexes. They follow the numbering in the original paper or database and often do not run continuously from 1. Residue 100 may be followed by 105.
Numbers may include letters. Residues 42 and 42A are different. The letter is an insertion code, used to add a residue while retaining the original numbering scheme.
Some residues are absent from the file. Regions not observed experimentally have no coordinates. They may exist in the sequence but be missing from the structure.
Thus, "residues 40 through 60" differs from "array positions 40 through 60." Nadoo Bio does not send display indexes directly as model positions. Saving a selection fixes the source-file hash, model/chain/residue identifiers, and the mapping between displayed and internal positions. Positions are recalculated from that same source at execution. If the original file is unchanged, the saved selection is retained.
Small-molecule input
SMILES represents atoms and bonds as one line of text. SDF stores atoms, bonds, coordinates, and properties.
A molecule is not chosen arbitrarily from a multi-molecule file. Choose the one to use from the searchable list. Your selection is recorded in preparation history and can be reopened later.
An SDF file does not necessarily contain a valid 3D structure. Coordinates may be absent or meaningless.
What file inspection checks
Uploaded files are checked as follows.
| Checked | Not checked |
|---|---|
| Syntax follows the format's rules | Chemical validity of the molecule |
| Records are correctly separated | The molecule's identity or type |
| Counts of atoms, bonds, and other totals are consistent | Suitability as input for a particular model |
| Coordinates are finite numbers | Permission to execute the tool |
Passing inspection means the file is readable, not that it is ready for a particular computation. Scientific suitability is the researcher's judgment; execution eligibility is checked separately.
There is also settings validation: file inspection checks the file itself, while validation checks the values passed to a tool. Settings validation is free and can be repeated before execution. Invalid values are reported with the affected field and reason.
Recheck preparation when the source changes
Prepared input is tied to the source file's content hash. Uploading different contents under the same name changes that hash, so previous chain or region selections require review. This prevents a replacement file from silently changing your selection.
Common issues
A multi-molecule file will not proceed. Select which molecule to use from the list.
The selected region is empty. Its residues may not have observed coordinates. Check the available coordinate range in the structure viewer.
The sequence exceeds the length limit. Limits differ by tool. Check the tool reference and, when appropriate, consider splitting by domain.
The file was saved but exceeds inspection limits. Storage and inspection limits differ. Check decompressed size and item counts, then prepare the required subset in a valid format.
The file passed inspection but cannot run. Inspection and execution eligibility are separate. Check tool support for the input format and your execution permissions.
Next steps
After preparing inputs, choose settings and create a plan. Execution and approval explains plans and review. Reading results explains how to interpret values after computation.