Sequencing reads

PlasmidPop opens the files a sequencing run gives you: AB1 files from a Sanger (capillary) sequencer, and FASTQ files of nanopore or Illumina reads. Both carry a quality for every base; an AB1 file also carries the trace, the four coloured signals the bases were called from. Some nanopore services deliver their consensus as an AB1 file too.

Opening a read

Open or drop the file like any other (see Files and storage). A read opens as an ordinary document: its bases, with the enzymes, ORFs, primers and everything else working on them. What it adds is kept with it, in this browser, like the rest of the document:

• The status bar says Read, 92% Q20+: the share of bases whose quality is 20 or more. Quality is on the Phred scale, where 20 means one error in a hundred bases, 30 one in a thousand. The threshold is the Align tab's Confident from (see Aligning a read with its qualities), so with Q40 set it reads Q40+.
• FASTQ qualities are read as Phred + 33, which every current sequencer writes. A file in one of the old encodings, Phred + 64 (Illumina 1.3–1.7) or Solexa, would read about 31 too high; when its quality characters look like one (none below @ and some above J, or a Solexa range from ; to past h) a warning says so, in the status bar or, for a file dropped into Align, in the note under the box. Convert such a file to Phred + 33 (for example with seqtk seq -Q64 -V) before relying on its qualities.
• A FASTQ file is read leniently where the meaning is clear, with a warning that counts what was changed: U in the bases is read as T, - and . as N, quality characters below ! as Q0, and the DEL character (0x7f) as Q93, the highest the encoding has. Any other character in the bases still refuses the file.
• An AB1 file keeps two copies of the base calls, as called and as edited. Qualities are used only from a copy of the same bases as the calls shown; if the edited calls differ from the original and have no qualities of their own, the read has none and the status bar says so.
• An AB1 read shows its trace in the sequence view, above the bases: the four dye signals (in the colours the bases are coloured in), each base's peak over its letter, and its quality as a faint bar behind. A clean read is a row of single, evenly spaced peaks; a mixed or noisy one shows overlapping peaks, which is where a base call is doubtful. The trace shows on a phone too. Format ▸ Trace draws it Short, Tall (twice the height, for peaks close together) or Hidden, for a read whose features and translations already make tall rows; the choice is remembered. It can be set while a read with a trace is in front.
• To check a read against a plasmid with its qualities, open the plasmid and drop the read on the Align tab's box: see Aligning a read with its qualities. Or, with the read open, drop the plasmid on the box: the read is aligned to it with its own qualities and trace, see When the document is the read.

A gzipped FASTQ (.fastq.gz), as nanopore runs usually come, opens without unpacking it first. A FASTQ file holds many reads and a tab holds one: the first read is opened, and the status bar's warnings say how many were left. To align another one, drop the file on the Align tab's box and pick the read there, or align every read in it at once with Align all (see Aligning a batch of reads), or assemble them into a consensus with no reference (see Assembling reads into a contig).

An AB1 file with no base calls (a fragment-analysis .fsa run) cannot be opened; there is no sequence in it.

Editing a read

A read's qualities and trace describe its bases as they came off the sequencer, and the trace is drawn only while they do. An edit that changes the bases — typing, deleting, pasting, moving the origin — sets them aside, and a notice says so; Undo brings them back with the bases. Edits that leave the bases alone (features, the name, the description) keep them. Reverse complement keeps them too, turned over with the bases.

Downloading a read

A download is GenBank, which has a place for the bases and features but not for qualities or a trace; a notice says so when you download a read.

File ▸ Export read as FASTQ writes the read's bases with their qualities, as a FASTQ file (Phred + 33, the encoding every current tool reads) named after the document, for an assembler, a variant caller or another lab. It is a download, like every export; the file you opened is not touched. The trace has no place in FASTQ and is left out, and so are features. The export is there only while the read is as it came off the sequencer: once an edit has changed the bases, the qualities no longer describe them, so the menu item is greyed out and says why. Undo back to before the edit and it is offered again. A document that was never a read has no such item.

File ▸ Export figure… with Sequence view (see Exporting) draws the trace as it is shown. The read stays whole in the browser. A share link carries the document the same way, so it too arrives without the read. To keep the original, keep the file you opened: PlasmidPop never writes to it.