#The three file types

.ab1, .fsa and .hid are one container, ABIF (see The format): a 128-byte header, a directory of tagged entries, and their data. What differs is the version, which tags an instrument writes, and in a .hid file what it encrypts. The deck reads all three with the same open; nothing depends on the file name.

The tag lists below are what the sample files in testdata/ hold (a 310 and a 3730 .ab1, a 3100 .fsa, a 3500 .hid). Other instruments and software versions add or drop tags, so test for a tag with has rather than from the extension.

#What all three carry

Every sample has the raw traces and the run's instrument record:

• DATA 1-4 (raw traces, one per dye), DATA 5-8 (electrophoresis voltage, current, laser power and oven temperature during the run)
• Dye# (the dye count) and DyeN (dye names)
• MCHN, MODL (instrument and model), TUBE (well), LANE
• RUND / RUNT 1-4 (run and collection start and stop)
• EVNT 1-4 (start and stop of run and collection), the settings EPVt (electrophoresis voltage), InSc and InVt (injection time and voltage), LsrP (laser power), Tmpr (run temperature) and Rate (scanning rate), Scan (the number of scans), SVER (software and firmware versions), MODF (data collection module), SMED and SMLt (polymer lot expiry and number), User

#.ab1: a Sanger trace

A sequencing run: four dyes, one per base, read in the order FWO_ gives (GATC). Besides the common tags it carries:

• DATA 9-12, the analysed traces, one per base
• PBAS, PCON, PLOC 1 and 2: bases, quality values and peak locations, as edited (1) and as called (2); read them with calls and editedCalls
• SMPL (sample name), PDMF (mobility file), S/N% (signal strength per dye), SPAC (peak spacing and basecaller), B1Pt (reference scan of the mobility curves), phAR, phCH, phDY, phQL, phTR (basecalling settings), RevC (whether the sequence is complemented)
• in the 310 file, MTXF / MTRX / MTRD (the spectral matrix), CMNT (a comment) and THUM (a thumb fingerprint); in the 3730 file, NOIS (noise per dye) and DyeW (wavelengths)

def f as abif.File init abif.open("sample.ab1");
def c as abif.Calls init abif.calls($f);
io.printf("%d bases, first ten %s\n", len($c.bases), $c.bases[..10]);

#.fsa: a fragment-analysis run

Sized DNA fragments (microsatellites, AFLP, STRs) in up to six dyes, one of them an internal size standard. The sample .fsa comes straight off the instrument:

• DATA 1-8 only: raw traces and the run's readings; the analysed traces (DATA 9-12) appear once analysis software has written them
• no base calls: calls refuses the file
• DyeN, DyeW (wavelengths) and DyeB per dye, DySN (dye set), GTyp (polymer), RunN, RMdN (run and run module), SpNm
• no SMPL: this file's sample name is empty in info

A fifth or sixth dye goes into DATA 105 and 106 (raw) and 205 and 206 (analysed); rawTrace($f, 5) finds them.

def f as abif.File init abif.open("run.fsa");
for (def d in abif.dyes($f)) {
    io.printf("%s: %d scans\n", $d.name, len(abif.rawTrace($f, $d.number)));
}

#.hid: a 3500 HID run

The 3500 series' human-identification format: ABIF version 300, six dyes (the sample is a GlobalFiler run, size standard GS600LIZ). It adds the results of the instrument's own sizing and the record of every consumable:

• DATA 1-4 and 105-106 (raw), 9-12 and 205-206 (analysed), DATA 5-8
• StdF (size standard), SzFt (the sizing fit), SMap (the size map), Peak 1-25 (peak tables), ScPa (the size caller's parameters), SamQ
• AsyN, AsyV, AsyC (assay name, version and contents), AVld
• consumables: anode buffer (ABED, ABID, ABLt, ABRn, ABTp), capillary array (CAED, CAID, CALt, CARn, CASN), cathode buffer (CBED, CBID, CBLt, CBRn, CBTp), polymer (SMID, SMRn)
• InjN (injection), SpeN (spectral calibration), RNmF (Raman normalisation), OvrI / OvrV (values over 32767 that the 16-bit trace could not hold), AmbT (ambient temperature), CkSm (checksum), AAct, AUDT (audit trail, XML)

Most of the measured data is stored as element types 28 to 34, which the published format does not define and which are encrypted: every DATA trace, the peak tables, the size map and fit, the assay contents, the saturation and offset scans, dye names 1 to 5, the dye set and the model. The deck returns them as Value.Raw, rawTrace refuses them, and info and dyes leave those names "". Sample name, well, instrument, run times, size standard, wavelengths and the consumable records read as usual.

def f as abif.File init abif.open("run.hid");
def i as abif.Info init abif.info($f);
io.printf("%s on %s, size standard %s\n", $i.sample, $i.instrument, $i.sizeStandard);

Reading the signal of a HID run needs Applied Biosystems' software.