Batch examples

These examples use One row per job. Add your own files to the matching inputs before importing, and download the selected tool's Template to check its current headers.

Boltz-2 complexes

A row can combine protein chains, ligands and other supported molecules. This Boltz-2 layout creates a two-chain complex with a ligand, then a protein with two copies:

csv

name,protein_1,protein_2,ligand_smi_1,protein_1.copies
complex-a,file:chain-a.fasta,file:chain-b.fasta,CCO,1
homodimer-b,file:chain-a.fasta,,,2

Each referenced FASTA contains one protein chain. The separators preserve absent optional inputs. Use separate DNA, RNA and CCD ligand columns for those molecule types; do not put them in protein or SMILES columns.

For a protein's precomputed alignment, add protein_1.msa with a value such as file:chain-a.a3m. Keep chain identifiers, templates and restraints consistent with the complex. The attribute reference lists supported controls.

BioPhi antibody pairs

For BioPhi Sapiens (Humanization) or Both (Humanization + Scoring), keep each intended pair on one row:

csv

name,heavy_chain_1,light_chain_1
antibody-a,file:vh-a.fasta,file:vl-a.fasta
heavy-only-b,file:vh-b.fasta,

Each file contains one complete antibody variable domain. The second job has no light-chain input. Avoid pairwise list mode here: it removes blanks and can reuse the single light chain.

For OASis (Humanness Scoring), use antibody_sequences_1 instead. Reference one FASTA containing the chain or paired VH/VL records that belong to that job. Switching operations changes the input columns.

PDBFixer with different settings

A file manifest is useful when structures need different treatment. This PDBFixer batch varies Add missing atoms:

csv

name,protein_1,setting.add_missing_atoms
repair-a,file:sample-a.pdb,true
inspect-b,file:sample-b.pdb,false

Other settings use the shared form values. For identical treatment of every file, regular bulk upload may be simpler. CSV does not change the tool's structure or model requirements.

MAFFT alignments

Each MAFFT job needs the sequences to align together:

csv

name,sequences_1
family-a,file:family-a.fasta
family-b,file:family-b.fasta

Each referenced file contains multiple sequences and remains one input to one alignment job. Placing every sequence in a separate job would ask the tool to perform separate analyses.