Sequence Analysis and BLAST
Everything that reads a sequence and tells you something about it: what it matches in the public databases, what it looks like on the other strand or in another frame, how it lines up against your other sequences, and what a protein or an RNA is likely doing structurally.
Two entry points. Tools that work on a stretch of bases live in the selection menu, covered in Plasmid Maps and the Sequence Viewer. Tools that work on the whole sequence live in the sidebar's Analysis section, and that section only shows what applies, so a protein sequence offers the protein tools and an RNA sequence offers RNA structure.
BLAST
BLAST in the selection menu searches your selection, or the whole sequence when nothing is selected, against NCBI.
Database offers Nucleotide (nt), Protein (nr), and SwissProt, starting on the one that matches your sequence. The Program below it is worked out from that choice and your sequence together: nt runs BLASTN whatever you give it, and nr or SwissProt run BLASTX on nucleotides and BLASTP on protein. Max Results takes 10 to 100 hits and starts at 50.
Hits come back as a sortable table of accession, name, organism, identity, E-value, bit score, and alignment length, sorted by identity. Open a row for the full name, the record's own length, and the raw score.
Saving a hit fetches the full record from its accession, so what lands in your library is the real annotated sequence rather than an alignment fragment. It is saved straight away, named for its accession and organism, and it records the sequence you searched from as its parent.
Sequence lookups leave Conspecta for NCBI and UniProt. That's deliberate and it's disclosed on every screen that does it. A BLAST search sends your whole query sequence through our servers to NCBI. Fetching a record by accession, searching by keyword, and saving a BLAST hit go straight from your browser, so those databases also see your IP address. Nothing else from your project goes with any of it, no account, project, or contact details. Avoid searching sequences you need to keep unpublished.
Transforming a Selection
The Transform group in the selection menu rewrites the bases you picked and shows you the result rather than changing your sequence. Reverse Complement, Complement, and Translate are there for DNA and RNA, Transcribe for DNA, and Reverse Transcribe for RNA.
Find ORFs, under Analyze, scans all six frames of your selection for open reading frames of 30 amino acids and up, whether or not they start at a start codon, and lists them longest first with their frame, strand, position, and whether each has a start and a stop. That is a looser search than the ORFs layer on the map, which only draws frames of 100 amino acids and up that begin at a start codon, so the tool will find things the map does not draw.
Every result numbers its positions against the whole sequence, not against your selection, so a position you read here is a position you can jump to.
A single-sequence result, such as a reverse complement or a translation, has Copy and Save as new sequence underneath it. Saving asks for a name, tells you it saved, and leaves you where you were with Open sequence if you want to go and look. The original is untouched, and the new sequence records where it came from.
Find ORFs and Translate Frames produce a list rather than one sequence, so each row carries its own copy button instead.
Aligning Sequences
Multiple Alignment in the sidebar lines this sequence up against others you pick. Only sequences of the same type are offered, and you need at least two. Conspecta builds a guide tree and merges the sequences along it, pairwise, which is the standard progressive approach.
The viewer breaks the alignment into blocks of sixty columns, each with its own position ruler, a conservation bar coloured from red to green, one row per sequence, and a consensus row underneath. Above the blocks it reports how many sequences and positions there are, and the average conservation across them. Hover any column of the bar to read what fraction of the sequences agree there.
Translate Frames shows all three forward reading frames of the whole sequence at once, each with its length and where it stops.
Trees, Dotplots, and Logos
Phylogenetic Tree builds a UPGMA tree from sequences you pick, and unlike the aligner it lets you mix types, so check what you selected before reading anything into the topology. Download SVG writes a real vector, branches as lines and tip labels as text you can edit in Illustrator, and Copy Newick takes the tree itself to the clipboard.
The Import Newick tab reads a tree someone sent you. Paste it in, or use Upload File to read a .nwk, .tree, .newick, or .txt into the box, then press Import Tree to draw it.
Dotplot compares this sequence against one other and draws where they match. Window Size runs from 5 to 25 bases and starts at 11, and Match Threshold runs from 50 to 100 percent and starts at 70. Underneath the plot you get the number of matches, how many diagonal runs there are, and the longest one, which together say whether the similarity is a real colinear stretch or scattered noise.
Sequence Logo stacks the conservation of several sequences into the familiar letter tower, with the y-axis in bits. It needs at least two sequences of the same type and the same length, since it reads column by column and does not align for you. Run Multiple Alignment first and use the result.
RNA and Protein Structure
RNA Structure, on RNA sequences, folds the sequence to maximise base pairs and draws the result as an arc diagram, as dot-bracket notation, and as a list of stems and loops. Min hairpin loop sets how tight a turn is allowed, from 1 to 5 bases, and starts at 3. Allow G-U wobble pairs is on. The Pairing score it reports counts pairs, so read it as a comparison between foldings of the same sequence rather than as a free energy.
Three tools read proteins:
| Tool | What it gives you |
|---|---|
| Hydropathy Plot | A Kyte-Doolittle trace over a window you set from 5 to 21 residues, the overall GRAVY value, and the stretches hydrophobic enough and long enough to be called as likely transmembrane regions |
| Protein Properties | Isoelectric point, net charge at pH 7.4, extinction coefficient at 280 nm both with cysteines reduced and with them paired, the matching A 0.1% values, the Trp, Tyr, and Cys counts behind them, and molecular weight |
| Secondary Structure | A Chou-Fasman prediction as helix, sheet, and coil percentages, a coloured map along the sequence, propensity plots, and each predicted region with its span and score |
A protein with no tryptophan, tyrosine, or cysteine says so rather than reporting an extinction coefficient, because there is nothing at 280 nm to measure.