Molecules

What it doesCopy Anchor Link

Opens a molecule file or a reaction as a real tab in the project, drawn as a structure with its properties beside it.

When you'd use thisCopy Anchor Link

The reagent, not the sequence. A construct carries an ampicillin-resistance gene; the thing that gene protects against is a molecule, and until now that molecule lived in a browser tab on somebody else's website. The same goes for the inducer you added, the substrate your reporter cleaves, the dye in your qPCR master mix, and the small molecule a screen came back with.

It is also where you go to check a structure rather than admire it: does this SDF actually contain the compound its filename claims, are the stereocentres specified or has the vendor sent a mixture, does this SMILES string from a supplementary table parse at all.

What this computesCopy Anchor Link

Two tables, deliberately kept apart: what was computed from the structure, and what the file merely said. A property a database wrote into an SDF tag is not a measurement this tool made, and mixing the two is how a wrong number acquires authority.

Computed from the structure: molecular formula, net charge, exact and average mass, cLogP, TPSA, hydrogen-bond donor and acceptor counts, ring count, rotatable bonds, stereocentres, the canonical SMILES, and the InChI with its key — the last two being what make one structure comparable with another across tools.

A property is defined against a form, and the form is a control. Hydrogens can be drawn as the file wrote them, all explicit, or implicit — and implicit is the default because the properties are defined against it. A record downloaded from a database usually writes its hydrogens out, and under explicit hydrogens the rotatable-bond count comes out higher, because an oxygen holding a hydrogen stops being a terminal atom. The panel does not hide that; it lets you switch and watch the number move.

Substructure search: Highlight SMARTS marks every match of a query on the drawing. A pattern the toolkit cannot parse says so; a valid one that matches nothing says no match — the two are different answers and are reported differently.

Per file: which record of many is drawn (an SDF or a SMILES file usually holds several), and the format recognised from the file's own contents rather than its name — an SDF is a file with $$$$ record separators, whatever it is called.

What it refuses, by name: ChemDraw CDXML, Ketcher KET and InChI as an input are not read and say so rather than opening empty, and a file over 64 MB is refused with its size stated.

Where it livesCopy Anchor Link

Tools → Analyze → Open Structure, or the command palette. A .mol, .sdf, .smi or .rxn also opens directly from the Files rail and the file manager, and previews in place there. A file picked from another project is copied into the open one so the tab survives a reload.

Inputs and outputsCopy Anchor Link

Takes MDL Molfile, SDF, SMILES and MDL reaction files, recognised from the file's own contents rather than its name — an SDF is a file with $$$$ record separators in it, whatever it is called. Files over 64 MB are refused with the size stated.

ChemDraw CDXML, Ketcher KET and InChI-as-an-input are not read, and say so rather than opening empty. An InChI and its key are still computed for anything that does open.

There is no output file: this is a reader.

How to read the resultCopy Anchor Link

One structure is drawn at a time, with the record number as a control — an SDF or a SMILES file usually holds many. Every choice about how it is drawn is a control rather than a default you cannot see:

• Hydrogens — as the file wrote them, all of them, or none. The default is implicit, and it is the one that matters: a record downloaded from a database writes its hydrogens out explicitly, and the properties beside the drawing are defined against the implicit form. Rotatable-bond counts in particular come out higher with explicit hydrogens, because an oxygen holding a hydrogen stops being a terminal atom.
• Kekulé — aromatic rings as alternating bonds instead of a circle.
• Stereo labels — R/S and E/Z annotations beside the wedges.
• Atom numbers, CoordGen layout and Rotate — for reading a fused ring system, or lining a drawing up against a published figure.
• Highlight SMARTS — mark every match of a substructure query on the drawing. A pattern the toolkit cannot parse says so; a valid one that matches nothing says no match.

Beside the drawing are two tables, deliberately separate: what was computed from the structure (formula, charge, masses, cLogP, TPSA, hydrogen-bond counts, rings, stereocentres, canonical SMILES, InChI and key) and what the file itself declares. Two readouts of the same quantity can legitimately differ — polar surface area is the usual case, because definitions disagree about whether sulfur counts — and merging them would hide which one you were reading. Full detail: the chemistry canvas.

Pitfalls worth knowingCopy Anchor Link

• A structure with unspecified stereocentres is a mixture, not a drawing problem. The stereocentre reading calls them out. Every property above it is then being reported for one of several molecules.
• A SMILES file has no signature. Any line of text could be one, so a .smi is routed by its extension alone — which also means a file full of SMILES saved as .txt will not open here.
• Polar surface area will not match your database. That is expected and the tooltip says which definition is in use; it is not a rounding difference.

RelatedCopy Anchor Link

Sits beside the table viewer, the NumPy array viewer and the PDF reader — the documents and data a project accumulates around the sequences it is really about.