Sequences

What the workbench is

The Sequences workbench is a SnapGene-style sequence editor built directly into ResearchOS. It holds every DNA, RNA, and protein sequence in your lab folder in one indexed library, and it opens those sequences in an editor that renders circular plasmid maps, annotated linear maps, and base-level sequence views without any external software. The workbench lives at /sequences and is always one click away from the rest of the app.

The point is that your sequences live alongside your experiments and notes, not as attachments tucked inside a separate application. A plasmid you designed last month, the oligos for a current cloning reaction, the protein from a collaborator's GenBank file. They all land in the same library, open in the same editor, and are searchable by the same search box.

The sequence library

The left panel of the workbench is the library. It lists every sequence in your folder, grouped into collections that map to your projects. A collection selector at the top of the panel lets you narrow the list to one project or view everything at once. Within a collection, sequences sort by name (ascending) by default, and the column headers let you re-sort by date added, type, or length. A search box at the top of the list filters by name in real time.

Each library row shows the sequence name, the sequence type (DNA, RNA, or protein), the length in base pairs or amino acids, and the topology for DNA sequences (circular or linear). Clicking a row opens the sequence in the editor. A small button at the top of the library panel opens the New Sequence dialog, and an Assemble button launches the Cloning Workspace for multi-fragment assembly.

Getting sequences into the library

There are three ways to add a sequence. The New Sequence dialog creates a sequence from scratch. You name it, pick a type and topology, and type or paste the bases directly. The Import path accepts SnapGene files (.dna, .prot), GenBank (.gb, .gbk), FASTA (.fa, .fasta, including multi-record FASTA where each record lands as its own sequence), and plain-sequence text files by dragging them onto the library panel or using the file picker. A paste-from-clipboard option in the New Sequence dialog accepts raw base strings and auto-detects the type from the characters. Importing a GenBank file preserves the feature annotations, qualifiers, and topology flags the file carries, so a fully-annotated plasmid map from Addgene or a collaborator arrives complete.

Sequences produced by the Cloning Workspace land in the library automatically when you save the cloning product. Primers you design in the editor are saved as primer_bind features on the sequence, not as separate library entries, so the oligo stays with the template it belongs to.

Download from NCBI

A Download from NCBI button in the library header opens a guided flow that pulls a sequence straight from NCBI into the active collection. You can fetch a gene by symbol plus organism (for example GAPDH in Homo sapiens), a genome or assembly, or any accession, whether a nuccore record like NM_002046 or a genome accession like GCF_000005845. The record is fetched directly from your browser against NCBI's public API, so the only thing that leaves your machine is the public identifier you ask for, the same privacy model as the Primer-BLAST handoff. Annotated records arrive with their features intact. Genome downloads are size-capped, and a genome larger than the cap is refused with a clear message rather than freezing the editor. The downloaded sequence lands as a fully parsed entry carrying a “From NCBI” provenance badge, with its accession kept linkable.

The tab bar and view modes

The bottom of the editor panel carries a tab bar with five tabs, Map, Sequence, Features, Primers, and History. These tabs are the main view switcher for the open sequence, modeled on SnapGene's bottom navigation. The Features and Primers tabs carry a count badge showing how many annotations or primer_bind features the sequence currently holds.

The Map tab renders the sequence as a circular plasmid map or a linear map, depending on the sequence topology and the topology chip in the display strip. The Sequence tab renders the bases at nucleotide resolution with the complement strand, translation tracks, and ruler. The Features tab opens a structured list of every annotation on the sequence, with controls for adding, editing, or deleting features. The Primers tab lists every primer_bind feature as an oligo entry with its Tm, GC%, and binding site. The History tab shows the edit and save history for the sequence.

The display (Show) strip

A thin horizontal row of pill chips, labeled Show, sits in the editor chrome alongside the tab bar. This is the display strip. Each chip is a toggle, and pressing it switches a display layer on or off without changing the active tab. The strip carries eight chips. Features toggles the whole annotation layer. Primers toggles primer-binding annotations. Enzyme sites overlays restriction-enzyme cut sites. Translation shows the amino-acid translation of CDS features. Open reading frames is a distinct toggle that highlights ATG-to-stop runs (over 30 aa, both strands) in unannotated DNA. Ruler / index shows the base-position ruler. Topology relabels in place (Circular for a plasmid, Linear for a linear molecule) and overrides a circular sequence to a linear layout. Wrap, active when the sequence is shown in linear form, switches between a wrapped multi-row layout and a single continuous line. On the Features, Primers, and History tabs (where there is no canvas to draw on) the chips dim and go non-interactive in place rather than disappearing, so nothing jumps as you switch tabs.

The Features chip carries a disclosure caret on its trailing edge. Clicking the caret opens a small flyout listing every feature type present on the sequence, with an eye icon per type so you can hide specific annotation categories (promoters, for example) without hiding everything. An amber dot on the Features chip marks when some types are hidden, so you never lose track of a non-default visibility state.

The circular and linear map

When the Map tab is active, plasmid sequences render as a circular map, a color-coded ring where each annotated feature appears as a named arc in the color assigned to its feature type. Clicking a feature arc on the map selects it and scrolls the Features panel to that entry. The topology chip in the display strip switches a circular sequence to a linear map if you prefer the linear layout, and a linear molecule always opens in linear form.

The base-level sequence view

The Sequence tab renders the sequence at nucleotide resolution. The forward strand runs left to right; a complement strand renders below it; feature and primer annotations appear as labeled colored tracks above or below the bases; and a coordinate ruler above the bases shows position numbers. At high zoom levels the individual characters are readable and editable; at lower zoom levels the view summarizes the sequence as a density overview. A pinch gesture or the zoom slider at the bottom of the editor adjusts the zoom level continuously.

Editing a sequence

The editor is editable by default. Typing at the cursor inserts bases; backspace and delete remove them; and selecting a range and typing replaces it. The keyboard shortcuts for copy, cut, paste, select-all, undo, and redo all work as they do in a text editor. Cut and paste carry the molecular context, so pasting over a selection updates any feature coordinates that span the replaced region, and pasting a copied range that included features proposes bringing those features along.

A toolbar above the editor carries the Undo, Redo, and Save buttons, plus dropdown menus for Edit operations, Feature and Primer actions, Enzyme management, and Analyze tools. Saving is explicit. The Save button records a new checkpoint in the sequence history. The History tab surfaces every saved checkpoint with a diff and an optional restore path.

A Find box in the toolbar searches by DNA sequence (exact match, with a closest-match fallback for near-misses), by feature name, or by protein sequence. Each match highlights in the viewer and the arrow keys step through multiple hits.

The live selection readout

When you click or drag to select bases in the editor, a readout bar at the bottom of the editor updates immediately with the selection coordinates, the length in base pairs, and the GC content as a percentage. For selections between 8 and 50 bp the readout also shows the predicted Tm, computed using the same nearest-neighbor thermodynamic model the primer tools use. A floating badge near the cursor shows the same information while you are still dragging, so you can see the stats before you release. Clicking a feature arc rather than dragging shows the feature name alongside its coordinates.

Annotating a sequence

Features are the annotation layer of a sequence. Each feature has a name, a type (CDS, promoter, primer_bind, rep_origin, and so on), a strand, a color, a set of coordinates, and an optional set of GenBank qualifiers (/product, /gene, /note, and others). The Sequences workbench provides several paths to add or update features.

The most direct path is to select a region on the map or sequence and open the feature editor from the context menu or the Feature toolbar menu. The feature editor dialog mirrors the SnapGene “Edit Feature” interface, with a name, type, strand, a segment table for multi-segment features, a color swatch, and the qualifiers editor. A “Translate in sequence view” toggle marks CDS features for translation display; a “Prioritize display” toggle keeps important features legible on dense maps.

For sequences without annotations, two automated paths under the Analyze menu can propose features. Detect Features runs a local database scan against bundled protein and DNA element reference sets. It identifies common elements such as fluorescent proteins, origins of replication, antibiotic resistance genes, and common promoters, and it reports each hit with its position, strand, and percent identity so you can review and accept or reject each proposal. Annotate from Reference aligns the open sequence against a second sequence you choose from the library and proposes to carry over the reference's features wherever alignment identity is high enough. Both paths land their accepted features in a single undoable edit.

Protein analysis

Selecting a CDS feature opens a protein inspector for the coding region. The inspector reads out the protein length in amino acids, the molecular mass in kDa, and the isoelectric point, and it offers three actions. Translate to protein turns on the amino-acid track on the map. Full protein properties opens a deeper view with the amino-acid composition, a hydropathy profile, and the extinction coefficient. Find domains runs an on-device HMMER scan in the protein panel, with an optional handoff to EBI InterProScan for a fuller search. Domain hits are labeled by source (EBI InterProScan, an on-device database, or a curated common-domains set) so you always know where a call came from.

A completed domain scan saves as a result artifact, so you can re-open it later from the History tab's Results section without recomputing it (see Sequence history below).

The cloning engine

The Cloning Workspace is a multi-fragment assembly tool launched from the Assemble button in the library header. It supports four assembly chemistries from one method-picker interface.

Overlap assembly (Gibson / NEBuilder HiFi) takes two or more fragments from the library or pasted directly, designs the junction overlap regions to a target length or Tm, and assembles them into a linear or circular product. The junction primers needed to amplify each fragment with the overlapping tails are computed automatically and offered as a copyable oligo order list or as primer_bind features on the product.

Restriction and ligation cuts two sequences at chosen restriction enzyme sites (from a set of common cutters, EcoRI, BamHI, HindIII, PstI, KpnI, SmaI, XhoI, NotI) and ligates the compatible ends, computing all possible directional products and letting you pick the intended one.

Golden Gate (Type IIS) uses a chosen Type IIS enzyme (BsaI, BsmBI, BbsI, or SapI) to assemble fragments via non-palindromic four-base overhangs, producing a scarless product at each junction. Gateway (BP or LR) simulates the recombination-based transfer between entry and destination vectors, computing the att site recombination and the resulting construct.

All four chemistries share the same two-step flow. Pick fragments and set options, then review the computed product. The review screen shows the assembled sequence with its features and any warnings (such as internal cut sites that would be severed by the chosen enzyme). Saving the product creates a new sequence in the active collection.

Primer design

The primer tools in the Sequences workbench are built around the binding site, not just the sequence. The Add Primer dialog (accessible from the Primer toolbar menu or by selecting a region and using the context menu) takes a primer sequence by typing, pasting, or seeding from the current selection. As you type, it shows the length in nucleotides, the GC content, the predicted Tm, the number of binding sites on the template, and a visual alignment of the primer against its annealing region, with the 5' tail dimmed and the annealing bases highlighted over the template. A reverse-complement toggle switches the primer orientation without retyping.

Double-clicking an existing primer_bind feature on the map or sequence opens the Edit Primer dialog, which shows the same stats and alignment but seeds everything from the stored feature. Editing the oligo in this dialog re-derives the binding site and Tm live; saving writes back to the feature. Primers can also be given an explicit color independent of the default primer type color, so forward and reverse primers on the same template are visually distinct.

The Primer Design mode in the Add Primer dialog includes a mutagenesis path. Opening the dialog in mutagenesis mode lets you specify a substitution, insertion, or deletion at a chosen position, and the engine designs a primer with the mutation incorporated and the flanking bases selected to hit a target Tm.

Specificity checking via NCBI

After designing a primer, a specificity check against NCBI BLAST is available from the Primers panel. Clicking the BLAST link for a primer opens the NCBI Primer-BLAST interface in a new tab with the primer sequence pre-filled, handing off to NCBI's servers to check for off-target binding against any genome in their database. This is a handoff, not an embedded result. The check runs in the browser against NCBI's public API, and the result stays in the NCBI tab.

Comparing and aligning sequences

The Compare / Align dialog (accessible from the Analyze menu in the editor or from the library header) accepts two sequences from the library and runs a pairwise alignment. Two alignment modes are available, global (Needleman-Wunsch, end-to-end alignment) and local (Smith-Waterman, best matching region). Scoring is IUPAC-aware for DNA, so degenerate base codes in either sequence match their constituent bases correctly, and a BLOSUM62 protein scoring mode is available for amino acid sequences. The dialog auto-detects whether the sequences look like protein or DNA from their character content and switches the default scoring accordingly.

The result shows the percent identity, the number of mismatches and gaps, and the full wrapped alignment with a match/mismatch midline and coordinate ticks. For shorter sequences a k-mer dotplot renders below the alignment, giving a visual read on repeat structure or structural similarity. Sequences longer than roughly 60,000 bases cannot be aligned end-to-end in the browser without hanging; the dialog reports this limit and suggests the local mode instead.

Mutagenesis

Site-directed mutagenesis primer design is available directly from the sequence editor. Select the target region, open the Add Primer dialog from the Primer menu, and switch to Mutagenesis mode. The mutagenesis designer accepts three mutation types, substitution (change one or more bases to specific alternatives), insertion (add bases at a position), and deletion (remove a span of bases). Given the mutation spec and the flanking template sequence, the engine computes a primer that incorporates the mutation with flanks long enough to hit a target Tm. The resulting primer can be added to the sequence as a primer_bind feature or copied directly to the clipboard for ordering.

Sequence history

Every explicit Save in the sequence editor records a checkpoint in the sequence history. The History tab shows these checkpoints as a timeline, with the date and time of each save. Selecting a checkpoint shows the sequence state at that point and the diff between that checkpoint and the current version. Restore from the history tab rolls the sequence back to the selected checkpoint in one click, creating a new checkpoint so the roll-forward is also preserved.

The History tab also carries a Results section above the version timeline. Operations that produce something, a completed Compare / Align run or a domain scan, persist there as re-openable result artifacts rather than throwaway popups. Re-opening one re-renders the stored result without recomputing it. Each artifact records a fingerprint of the sequence at the moment it was computed, so a result picks up a staleness flag once the sequence has changed since the run, telling you the stored result no longer reflects the current molecule.

Exporting sequences

Sequences can be exported in GenBank format or FASTA format from the Export menu in the editor toolbar. The GenBank export preserves all features, qualifiers, and topology flags in the standard .gb format that any sequence analysis tool can read. The FASTA export writes the raw sequence string with the sequence name as the header. A selection-to-FASTA option exports only the selected region, and a selection-to-protein-FASTA option exports the translated amino acid sequence of the selected CDS. The circular map can also be exported as a PNG or SVG image from the map view. A saved sequence map also drops into the Figure Composer at /figures as a per-panel-styleable panel, so you can lay it out alongside plots, trees, and molecules in a multi-panel publication figure and export the whole page as one clean vector SVG.

Connection to the rest of the app

Sequences participate in the shared folder structure alongside notes, experiments, and methods. A sequence is filed under a project through the collection it belongs to, so it shows up in the same project context as the experiments that use it. Deleting a sequence goes through the same trash flow as every other record type, with the same 30-day recovery window and the same restore behavior.

A sequence also embeds as a live read-only block inside a note, rendered either as a map ribbon or as a bases preview, so a write-up can show the construct without leaving the editor. Enriching a sequence's organism from NCBI fills in its taxonomic lineage, and an “Explore in tree” link then opens the tree view centered on that organism.

BeakerBot can drive the workbench too. The assistant can compute Tm, translate a sequence, take a reverse complement, find open reading frames, design primers, fetch a record from NCBI, and run Gibson or digest-and-ligate assembly, all through the same engines the editor uses.

The primer Tm engine in the Sequences workbench is the same nearest-neighbor model used elsewhere in the app. The same nearest-neighbor Tm model powers the Primer Tm tab in the Lab calculators modal, so a Tm you compute from a selection in the sequence editor and one you compute in the calculator will always agree.