Molecule World User Guide: Apple 4.6
A molecular viewer for iPhone, iPad, and Mac. Open a structure from a file or fetch one by name, then explore it in 3D with multiple display styles, color schemes, and selection filters.
Version 4.6
Contents
1. Getting a structure on screen
You can load structures from local files or database searches.
Open a local file
Everything you open or fetch collects in the app's library list, with newly downloaded structures added at the top. The app reopens your last-used structure when you return.
If iCloud Drive is on, your library — and your custom ordering — syncs automatically across your iPhone, iPad, and Mac. If iCloud is off, everything still works; your structures are simply kept on the device.
To manage the list, enter Edit mode:
- Drag to reorder — set your own order; it syncs across your devices.
- Swipe to delete a single structure, or select several and delete them together.
- Filter by name to find a structure quickly in a long list. Your search term stays in place as you browse — it clears only when you change it, clear it yourself, or quit and relaunch the app.
Keyboard navigation (Mac & iPad)
With a hardware keyboard attached, you can move through the file list without touching the screen:
| Key | Action |
|---|---|
| Up / Down arrow | Move the selection up or down the list |
| Shift (Mac) | Open the selected structure |
| Return (iPad) | Open the selected structure |
This works the same way whether you're browsing the full list or a filtered (searched) one.
Folders and counts
Structures can be organized into folders. Each folder shows the number of structures it contains next to its name — for example Antibodies (23) — counted recursively, so structures in nested subfolders are included in the total.
The count bar at the top of the list is folder-aware. At the top level it shows the size of your whole library; open a folder and it shows just that folder's structures. While you're filtering by name, it shows how many structures match out of the total.
Drag and drop (Mac)
On Mac you can drag a structure file straight from Finder onto the 3D canvas to open it — no need to go through the library list first. Accepted types are PDB, CIF, BCIF, MMDB JSON, and CN3D (.cn3), including gzipped (.gz) and zipped (.zip) versions.
Windows and tabs (Mac)
On Mac you can keep more than one structure open at once. Right-click (Control-click) a structure in a folder and choose to open it in a new window or a new tab — handy for comparing structures side by side. Standard macOS window and tab controls apply, so you can rearrange, combine, or tear off tabs as usual.
Fetch by name or ID
Tap the plus button to open the search window. Pick a source with the segmented control, type, and search with PDB ID's (e.g. 1ID3) or keywords (e.g. hemoglobin, tRNA, DNA, antibody), or, in the case of PubChem, SDF ID's or compound names (e.g. aspirin, caffeine).
- RCSB PDB — Assembled structures in mmCIF(.bcif) format
- MMDB — NCBI structures with extra annotations in JSON format
- PubChem — Chemical structures in SDF format.
Each source shows a short description before you search, so you know what it's best for, and results replace it once the search runs. Results come back as a list. Tap one and it downloads, imports into your library, and loads. Large mmCIF-only entries fetch automatically as a BCIF assembly, so big complexes still open.
Searching PubChem
PubChem search understands more than compound names — it runs against PubChem's quick-search back end, so you can search by:
- Name or synonym — e.g. aspirin, caffeine (CAS numbers are indexed as synonyms, so they work too).
- Molecular formula — e.g.
C9H8O4. Molecule World recognizes a formula and matches it as a whole, rather than splitting it on its digits. - SMILES — paste a SMILES string (e.g.
CC(=O)Oc1ccccc1C(=O)O) and the app runs a structure-identity lookup, falling back to a text search if there's no exact match.
Multi-word queries are combined with an implied AND between words, and you can also use quotes, wildcards (gluc*), OR / NOT, and field tags such as aspirin[CompoundName] or 180.16[MolecularWeight]. Molecule World detects SMILES and formulas for you, so you don't have to tag them yourself.
File Types
Molecule World works with several file types listed in the following table.
| Format | Typical source | Notes |
|---|---|---|
PDB (.pdb) |
RCSB, hand-built files | Plain-text protein / nucleic acid structures |
CIF (.cif) |
hand-build file | Plain-text protein / nucleic adic structures |
BCIF (.bcif) |
RCSB ModelServer | Biological and large structure assemblies |
MMDB JSON (.json) |
NCBI | Richest annotations (domains, etc.) |
CN3D (.cn3) |
NCBI legacy | ASN.1 structures |
SDF (.sdf) |
PubChem | Small chemicals |
Any of these can be dropped onto the Mac canvas, and PDB, CIF, BCIF, JSON, and CN3D files are also accepted gzipped or zipped.
2. Moving around in 3D
Molecule World uses standard iOS and trackpad gestures to manipulate structures.
| Gesture | Action |
|---|---|
| One-finger drag | Rotate the structure |
| Pinch | Zoom in / out |
| Two-finger drag | Pan |
| Auto-rotate (spinner) | Spin the structure hands-free |
On Mac, you can use either the trackpad gestures above or a mouse:
| Mouse | Action |
|---|---|
| Left-click drag | Rotate the structure |
| Scroll wheel | Zoom in / out |
| Option (⌥) + left-click drag | Pan |
Keyboard controls (Mac & iPad)
With a hardware keyboard attached, on Mac or iPad, you can drive the view from the keyboard. Hold a key to keep moving, and press two arrows together to combine directions (e.g. diagonal panning).
| Key | Action |
|---|---|
| Arrow keys | Rotate the structure |
| Option (⌥) + Arrow keys | Pan |
| Command (⌘) + / Command (⌘) − | Zoom in / out |
MoleculeWorld remembers how you left each structure — style, colors, selection, spin, and camera are saved per structure and restored next time you open it. Tap reset (scope) to return to first-load defaults: the format-default style, all atoms shown, and centered framing.
3. Display styles
Switch styles from the viewer controls. Available representations:
- Spacefill — atoms at full van der Waals radius, bonds hidden. Shows overall shape and packing.
- Ball-and-stick — atoms as spheres, bonds as cylinders. The general-purpose view.
- Tubes (licorice) — thin bonds only, no large joint spheres. Good for tracing connectivity and backbones.
- Ribbon (cartoon) — smooth secondary-structure ribbons: helices as rounded coils, β-strands as arrows. The standard protein view.
- VdW surface — a molecular surface skin over the structure.
- Wire surface — the same molecular surface, drawn as a wireframe mesh instead of a solid skin, so you can see the surface shape while still seeing what's underneath.
Heterogens (ligands, ions, cofactors) follow special rules so they always read clearly: organic heterogens render as element-colored sticks/tubes, and metals render as element-colored spheres at covalent radius — in every style, including ribbon and surface.
4. Color schemes
Coloring is independent of display style — pick any combination. Switching a color scheme is fast and doesn't reload the structure.
| Scheme | What it shows |
|---|---|
| Element (CPK) | Standard per-atom element colors (C, N, O, S…) |
| Secondary structure | Helix / sheet / coil |
| Chain | Each chain a distinct color |
| Domain | Named structural domains — see section 7 |
| Residue type | RasMol / Shapely convention (amino acid types and purines and pyrimidines for nucleic acids) |
| Residue | Each amino acid or nucleotide has a specific color |
| Charge | Positive / negative / neutral (nucleic backbones read negative), shown at pH 7 |
| Hydrophobicity (DWB) | Uses a hydrophobicity palette |
| B-factor | Blue-white-red gradient scaled to the structure's own B-factor range, highlighting areas of higher atomic motion or positional uncertainty |
| Rainbow | Gradient along the sequence (N→C terminus) |
In chain, rainbow, and other residue-level schemes, heterogens stay colored by element so ligands and ions remain identifiable. The legend notes this where it applies.
B-factor coloring is available for PDB, mmCIF/BCIF, and MMDB structures that carry B-factor (temperature factor) data; it's dimmed for structures that don't — for example CN3D files, or predicted structures with no real per-atom variation to show.
5. Showing part of a structure
Use the Show selector to focus on a subset of molecules, or backbone representation, in the structure. Everything outside the selection is hidden.
- All — the whole structure
- All + water — the whole structure, plus crystallographic waters
- Protein — amino-acid chains only
- Nucleic acids — DNA / RNA only
- Heterogens — ligands, ions, cofactors
- Core backbone — backbone trace
- Complete backbone — full backbone
- Cα trace — alpha-carbon trace
Water. Crystallographic water molecules are hidden by default — in most views they obscure the structure more than they explain it. Choose All + water to show them as red oxygen atoms; they appear in every display style, including ribbon. Water shows only with this option, not with the other filters, and your choice is remembered per structure.
Dimmed options. Color schemes and Show items that don't apply to the loaded structure are dimmed and can't be selected — only the combinations that make sense for what you've opened stay active. For example, Domain is dimmed when no domains could be identified, All + water when the structure contains no water, Secondary structure when no helices or sheets are assigned, Nucleic acids when the structure contains none, and B-factor when the structure carries no usable B-factor data.
6. Sequence view (iPad & Mac)
On iPad and Mac, a Sequence bar sits at the bottom of the canvas next to Show. Tap Sequence — its label shows the chain count — to slide up a panel with one row per chain: protein chains first, then nucleic acid chains, then heterogens.
Sequences render as single-letter codes and are colored to match whatever color scheme is active, so switching to Domain coloring shows the domain segments right in the sequence. Heterogens are the exception — they display as short codes (typically three letters, or a condensed symbol for a modified residue) rather than a single letter, since there's no standard one-letter code for most ligands and modified residues. Lone metal ions read with proper element casing (Cl, Cd, Zn) rather than the all-caps code stored in the file.
Modified nucleotides from older files — for example a methylated cytosine stored under a long chemical name — are condensed to their standard short symbol (e.g. m5C) instead of showing the raw name inline.
Use the Show button at the bottom to filter the sequence panel to just protein, nucleic acid, or heterogen chains, and drag the resize handle at the top of the panel to see more or fewer rows at once. Water is never listed in the sequence panel, even when it's displayed in the 3D view.
Keyboard and mouse in the sequence panel (Mac)
With focus in the sequence panel:
| Input | Action |
|---|---|
| Left / Right arrow | Step through residues |
| Up / Down arrow | Move between chains |
| Click-drag | Pan the sequence with the mouse |
| Scroll wheel / trackpad | Scroll the sequence |
Click or drag inside the sequence panel to give it keyboard focus; clicking back on the 3D canvas returns arrow-key control to the molecule view.
7. Color by Domain
The Domain color option shades each structural domain along a light-to-dark ramp within its chain, so the domain architecture is visible at a glance. It works with any structure format — PDB, CIF, BCIF, MMDB, and CN3D.
Domain names come from NCBI's CD-Search service, which identifies conserved domains from each protein chain's sequence. The lookup runs in the background when you load a structure; the picker and legend show progress, then label each domain by name. An antibody resolves into VH, CH1, CH2, CH3 and Vκ/Cκ; other proteins are named from the Conserved Domain Database.
The first lookup for a given structure needs an internet connection and can take up to a minute or two. Results are then cached per PDB ID, so reopening that structure is instant and works offline. A re-fetch control in the legend refreshes them.
Structures built from many copies of the same chain — viral capsids, for instance — are handled efficiently: identical sequences are searched once and the result applied to every copy.
Domain is dimmed when no domains could be identified — a structure with no protein chains, or one whose sequences match nothing in the database.
8. The color key
A legend panel (color palette button) shows what the current (selected) colors mean. It's filtered to what's actually visible — only the residues, elements, and heterogens present under your current Show selection appear, and categories like secondary structure or nucleic-acid entries only show up when the structure actually contains them, so the key stays relevant and sized to fit.
For heterogens the legend splits metals and non-metals, uses standard SNFG colors for sugars and Modomics-style symbols for modified nucleotides, and falls back to a CPK "heterogens by element" key in the element-colored modes.
In Domain coloring the legend lists each distinct color once rather than one row per chain, so a viral capsid with hundreds of copies of a chain gives a compact key instead of hundreds of near-identical rows. Crystallographic water is never listed.
9. Structure info
An info dialog ("I" button) shows metadata for the loaded structure — title, ID, and related details pulled from the file.
10. Tips & behavior
- Large structures may pause briefly while loading or when switching to ribbon/surface; a progress bar appears for longer operations.
- Per-structure memory restores your last view automatically — use reset (scope) to return to the default display.
- Surface and ribbon recoloring is quick because the geometry is cached; only the colors change.
- Tap the canvas to dismiss — on iPad and Mac, tapping the 3D view closes any open panel (Style, Color, Info, Color key), the Show selector, and the files drawer, all in one tap.
- About Molecule World (Mac) — choose Molecule World ▸ About Molecule World from the menu bar to see the app name, version, build number, and credits.
Appendix — Power user notes
For users who want to know why the viewer draws what it draws, and how to get the most out of the data.
A. How heterogens and metals are drawn
Ligands, ions, and cofactors are handled separately from the polymer so they always stay legible:
- Organic heterogens are colored by element in every scheme and drawn as sticks/tubes — including in ribbon and surface modes, where the main chain is a ribbon and the ligand still appears as connected atoms.
- Metals are colored by element and drawn as spheres at covalent radius (not van der Waals), so a lone ion reads as a compact ball. Metalloids are grouped with metals so they stand out.
This is why a ligand keeps its CPK colors even when the protein is colored by chain, rainbow, or hydrophobicity.
B. Bonds and connectivity
- Coordination bonds (metal–ligand contacts) render as dashed lines; bonds within a single residue stay solid.
- Disulfides are drawn with short side-chain stubs anchored to the backbone so the S–S link is visible in ribbon mode.
C. Color by Domain, in detail
Two independent sources feed this mode:
- Named domains from NCBI CD-Search (CDD) — available for every format, because they're derived from each chain's sequence rather than from anything stored in the file. Fetched in the background at load, then cached.
- Geometric domains from MMDB/JSON — NCBI's decomposition of the 3D structure into compact units. Available immediately and offline, with no lookup, but only for that format.
Antibody naming. When a chain's variable domain is recognized, its domains are named by convention — heavy chains as VH, CH1, CH2, CH3, light chains as Vκ/Cκ or Vλ/Cλ. A fragment with no variable domain (an isolated Fc, say) is labeled with its Conserved Domain Database name instead.
D. Selection semantics
Show is show-only: anything outside the selection is hidden, not greyed. The selection survives style and color changes, and the color key filters to match — only what's visible appears in the legend. Backbone tiers differ in scope (core vs complete), and Cα trace reduces a protein to its alpha-carbon path. Water is the one exception to the tiers: it's an overlay on All rather than a filter of its own.
E. Large structures and level of detail
Very large structures degrade gracefully rather than stalling:
- Above roughly 400K atoms, ribbon mode drops to a coil trace with reduced tessellation (large memory savings, confirmed near 800K atoms). The trace keeps its secondary-structure coloring, so helices and strands stay distinguishable even at reduced detail.
- On the largest structures a load may briefly block while geometry builds; this is intentional (an honest pause beats a fake-smooth delay). A progress bar shows for longer operations.
F. Surfaces
The VdW surface is a marching-cubes mesh, shared by both the solid and wire surface styles. Recoloring a surface is fast because the mesh is cached and only the colors are recomputed, and switching between solid and wire surface reuses the same cached mesh. (Masking the surface to a sub-selection currently shows the whole structure.)
G. Where structures come from
Fetched files carry their provenance in the filename (e.g. 1ID3.rcsb.bcif). Sources:
- RCSB — BCIF preferred, via the ModelServer assembly endpoint, provides biological assembly with full metadata.
- MMDB — JSON (domain annotations) or BCIF for very large structures.
- PubChem — SDF for small molecules, fetched through PubChem's quick-search back end so name, molecular-formula, and SMILES queries all resolve.
Large mmCIF-only entries are fetched as BCIF assemblies automatically.
H. Storage, iCloud, and view state
When iCloud Drive is enabled, your library is stored in iCloud Documents and kept in sync across your devices; structures you add on one device appear on the others as they sync. When iCloud is off — disabled, on a managed/enterprise device, or by choice — the app falls back to on-device storage with no loss of function.
Your custom file order is held on the device (so it works offline and for non-iCloud accounts) and mirrored to iCloud so it stays consistent across your devices. A couple of caveats worth knowing: deletions made while iCloud is off don't propagate to iCloud, and with iCloud off the library shows the structures that ship with the app rather than your synced set. A Restore sample structures option brings the shipped samples back if you've emptied the list.
Separately, each structure's full view — style, colors, selection, spin, water, and camera — is saved and restored per structure; reset (R) returns it to first-load defaults.
J. Snapshots
The snapshot/export path renders a high-resolution offscreen image with a transparent background (cropped to the structure) and opens the standard share sheet, so you can save it to Photos, save to Files, AirDrop it, or send it directly from Mail or Messages.
K. Color scheme reference
- DWB palettes are MoleculeWorld's own residue (nucleic acid bases) and hydrophobicity colors.
- Shapely is the RasMol residue-coloring convention used for animo acid residue coloring.
- Charge treats nucleic-acid backbones as net-negative, and reflects protonation at pH 7 — histidine in particular is only partly protonated near neutral pH, and the acidic/basic assignments shift at the extremes.
- B-factor coloring in MMDB JSON files autoscales to each structure's own range rather than a fixed scale, so "red" on one structure and "red" on another don't necessarily mean the same absolute value.
Abbreviations
The above text uses several abbreviations.
| Term | Stands for |
|---|---|
| ASN.1 | Abstract Syntax Notation One — the encoding behind CN3D files |
| B-factor | Temperature factor — a per-atom measure of positional uncertainty or thermal motion in a structure |
| BCIF | BinaryCIF — a compact binary encoding of CIF (below) |
| Cα | Alpha carbon — the backbone carbon of an amino acid |
| CD-Search / CDD | Conserved Domain Database — NCBI's domain-identification service |
| CIF / mmCIF | Crystallographic Information File / macromolecular CIF — the modern text format for structures |
| CN3D | NCBI's legacy 3D structure format ("see in 3D") |
| CPK | Corey–Pauling–Koltun — the standard per-element atom coloring |
| DWB | Digital World Biology |
| MMDB | Molecular Modeling Database — NCBI's structure database |
| PDB | Protein Data Bank — A repository of 3D structures |
| RCSB | Research Collaboratory for Structural Bioinformatics — operates the PDB |
| SDF | Structure-Data File — a small-molecule format (used by PubChem) |
| SMILES | Simplified Molecular-Input Line-Entry System — a text notation for a chemical structure |
| SNFG | Symbol Nomenclature for Glycans — the standard coloring for sugars |
| VdW | van der Waals — the atomic radius used for spacefill and surfaces |