VERSION	DATE		COMMENTS
-------	----		--------
1.0	2003-06-09	First official version

1.1	2003-06-09	tMin..tMax defaults to 0..100

1.2	2003-06-25	ct-energy reads .ct files with multiple structures
			hybrid-min & hybrid-ss-min produce valid .ct files when no structure is possible
			ct-energy parses empty structures correctly
			Makefile dependencies updated

1.3	2003-07-03	bug fixed: tMax defaults to 100 in hybrid2.pl, hybrid2-min.pl
			--noisolate in hybrid, hybrid-min, hybrid2-min.pl and hybrid-ss-noml
			hybrid and hybrid-min use less memory with --energyOnly
			--enthalpy and --entropy options in concentration(-same) and ensemble-dg(-same)
			compiler warnings of uninitialized variables eliminated
			command line arguments standardized across all programs
			hybrid2.pl and hybrid2-min.pl plot Au and Bu instead of A and B

1.4	2003-07-08	--noisolate in hybrid-ss-simple, hybrid-ss, hybrid2.pl
			bug fixed: P(i and i+1 SS) corrected in hybrid-ss, hybrid-ss-simple, hybrid-ss-noml
			compiler warnings about %lg in scanf() eliminated
			bug fixed: hybrid-ss-min traces bulge/interior loops correctly
			bug fixed: ct-energy computes exterior dangles of dimers correctly

1.5	2003-07-28	--mfold in hybrid-min
                        ct-energy displays components of loop energies with two --verbose's
                        probability cutoff of 1/1000000 in partition function programs
                        output of ensemble-ext, ensemble-ext-same normalized by sum of concentrations
                        output of ensemble-dg-same is scaled correctly (was 1/2 of what it should be)
			concentration and extinction plots have correct temperature range in hybrid2.pl, hybrid2-min.pl
			Cp2heat and dG2dH are file-based instead of stream-based
			Cp2heat writes multiple Tm's to TmCp
			dG2dH produces .H and .dH files
			dG2dS added to package
			ensemble-ext(-same) produces TmExt
			excluding AB disallowed
			--tracebacks option in hybrid-ss-noml, hybrid-ss-simple, hybrid-ss
			--quiet option in hybrid-min, hybrid-ss-min

1.6	2003-10-18	--tracebacks option in hybrid
			ensemble-ext(-same) determines extinction value correctly in TmExt
			hybrid2(-min).pl use strict, warnings
			extinctions added to hybrid-min, hybrid-ss-min, hybrid2-min.pl
			ensemble-ext(-same) produces TmExt1 (max derivative) and TmExt2 (halfway point)
			hybrid2(-min).pl understands --parallel when sequences are the same
			--dGscale option eliminated
			hybrid-ss-min, hybrid-min accept multiple sequences delimited by ;
			machine-readable energy rules renamed to *._[GH], *._[GH]D for Windows compatibility

1.6a	2003-10-24	machine-readable energy rules changed to *_.D[GH], *_.D[GH]D

1.6b	2003-10-27	machine-readable energy rules changed back to *.D[GH], *.D[GH]D
			human-readable energy rules moved to Rules subdirectory

1.6c	2003-10-31	bug fixed: --quiet works in hybrid-min

1.7	2003-11-06	extinctions corrected for unfolded strands
			{0,1,2,3,4} read as {A,C,G,T/U,N}

1.8	2004-02-04	prefilter and postfilter can be disabled separately
			dangling Ns have energy +infinity
			extra RT ln 2 term added to free energies of homodimers in hybrid(-min) and ct-energy
			searching for bulge/interior loops is faster
			--Vienna option in hybrid, hybrid-min and ct-energy
			significant speedups in hybrid-min, hybrid-ss-min

1.9	2004-04-27	--maxbp option in hybrid-ss, hybrid-ss-simple, hybrid-ss-noml, hybrid-ss-min, hybrid2.pl
			--force option in all partition function and energy minimzation programs
			--tracebacks works with --noisolate in hybrid, hybrid-ss, hybrid-ss-simple, hybrid-ss-noml
			--mfold in hybrid-ss-min
			temperature removed from names of .ct, .asc, .edp, .ann files if tMin + tInc > tMax
			all programs read FASTA format in addition to raw text

1.9b	2004-07-14	removed dependency on -liberty (helps with FreeBSD and MacOSX compatibility)

2.0	2004-11-08	1x1 loops with Watson-Crick or wobble mismatches have energy +infinity
			2x2 loops with 1 or 2 Watson-Crick or wobble mismatches have energy +infinity
			extra RT ln 2 term removed from free energies of homodimers
			individual heat capacities removed
			--Vienna option in hybrid-ss, hybrid-ss-noml
			hybrid-ss, hybrid-ss-simple and hybrid-ss-noml use less memory
			energy minimization programs do traceback correctly when dG = 0
			man pages written for each program
			autoconf/automake used to make building on various platforms easier
			prefilter can filter for helices of any length
			Postscripts from hybrid-plot-ng are in color and generally better
			hybrid2.pl uses gnuplot's enhanced postscript mode (subscripts, Greek deltas, etc.)
			--basepairs option in all PF and EM programs

2.1	2005-03-26	all programs read name from FASTA format
			--single option works in ensemble-ext-same
			extended .ct file format (two columns for stacking)
			tstackm and tstacke replace double dangle
			NUL, NULD and NULDAT rules created
			various changes to increase numerical stability
			concentrations(-same).pl normalizes so sum = 1
			-z/--zip option in hybrid-min and PF programs replaces -V/--vienna
			--title option in hybrid2.pl
			added support for hexaloops (special hairpin loops of size 6)
			HYBRIDDAT environment variable replaces HYBRIDLIB
			several bugs fixed

2.2	2005-04-08	1x2 and other interior loops with Watson-Crick mismatches have energy +infinity
			ensemble-ext(-same) normalizes properly
			ct-ext and ct-uniq added to package
			-s/-ss option added to ensemble-ext(-same)

2.2a	2005-05-19	plots are produced even when TmConc is undefined

2.3	2005-05-25	hybrid(-ss)-2s.pl, hybrid2-2s.pl added to package
			hybrid2.pl only calls gnuplot if it was found by configure
			ct-prob added to package
			it's back! RT ln 2 term added (again) for homodimers
			bug fixed in hybrid-min (multiple pairs of sequences)

2.4	2005-09-15	-q/--quiet works in hybrid-min
			TmExt2 is halfway between observed min and /theoretical/ max
			dG2d[HS], Cp2heat do not insist on .dG suffix
			hybrid-min output includes enthalpy and entropy with -q but not -E
			dHss based on enthalpy of strand and r.c., not ensemble
			TmConc defined by concentration of *unfolded* single strands
			fixed bug in extrapolation of bulge/interior and hairpin loops >= 30 bases
			prefilter now requires m of n possible pairs
			melt.pl added to package

2.5	2005-10-27	hybrid-plot-ng uses gd for gif/jpeg/png iff present
			ensemble-ext-same works correctly
			concentrations(-same).pl handled degenerate cases
			EM programs work correctly on platforms that don't understand 'inf'

3.0	2006-03-10	name changed to UNAFold
			hybrid-min works correctly if no basepairs are possible
			-C/--circular option added to hybrid-ss-min
			ct-energy correctly parses circular molecules
			--mfold works with multiple sequences in hybrid-ss-min
			ensemble-dg normalizes by max(A0, B0); ensemble-dg-same by A0
			plot2ann.pl, ss-count.pl added to package
			modifications to energy rules simplified: sint2.*, sint4.* and stack.DAT only
			partition function programs are ~70% faster
			hybrid-select.pl added to package
			h-num.pl added to package
			ct2rnaml added to package
			miscloop.* includes temperature
			scaling added to partition function programs
			--stream option added to hybrid(-ss)-min
			--help and --version work in (almost) all programs

3.1	2006-04-19	sequence files may be in different directories
			hybrid-plot compiles on MacOSX (configure --with-apple-opengl-framework)
			bug fixed in hybrid-ss --tracebacks
			ct-prob works correctly
			bug fixed: hybrid(-ss)-min --mfold works correctly
			hybrid-2s.pl works correctly when sequences can't hybridize
			ITC-plot.pl, vantHoff-plot.pl added to package

3.2	2006-05-25	memory error in ensemble-dg fixed
			ct-energy reports +infinity for hairpins of < 3 bases
			Cp2heat renamed to dG2Cp
			--hyperbolic option added to ITC-plot.pl
			configure accepts --disable-force
			standard output flushed in quiet mode after each output

3.3	2006-08-21	UNAFold.pl supports --model=PF
			--fraction works correctly on RNA in hybrid2.pl
			eliminated warnings in hybrid2.pl for undefined Tm's
			ct-uniq works correctly
			--NA works correctly in UNAFold.pl

3.4	2007-04-25	UNAFold.pl does hybridizations too
			salt correction added to tstackm and tstacke
			loops of > 30 bases treated correctly with .DH[D] rules
			eliminated warnings in plot2ann.pl
			sequence names > 80 characters don't cause crashes
			--constraints works in hybrid(-ss)-min
			infinite loops with stochastic tracebacks fixed
			melt.pl correctly processes options like --nodangle

3.5	2008-02-14	ct2rnaml outputs stacking information
			hybrid2.pl uses its name only (not its path) to determine special functionality
			eliminated an occasional warning in plot2ann.pl
			fixed bug involving plotting observed data in hybrid2.pl
			plot2ann.pl now handles mfold-style .plot files
			added support for prohibit-range constraints of the form "R i-j k-l"
			bug fixed: $UNAFOLDDAT is searched for extinction file
			new salt correction in polymer mode
			eliminated a warning in UNAFold.pl

3.6	2008-06-22	PF and EM programs check tMin/tInc for infinite loops
			many minor changes to man pages
			eliminated more warnings in hybrid2.pl for undefined Tm's
			better support for files in other directories, spaces in paths, etc.
			absorbances are calculated correctly for unknown bases
			UNAFold.pl outputs valid XHTML 1.0 Strict
			Bug fixed: energy details in UNAFold.pl work again
			UNAFold.pl supports -C/--Ct for dimers
			Bug fixed: hybrid-min (and thus melt.pl) calculates enthalpies correctly

3.7	2009-07-30	Better error messages for unopenable files
			Bug fixed: prohibit constraints in hybrid-ss-min
			Big fixed: asymmetric interior loops in PF programs
			2x3 loops are now calculated using tstacki23 tables
			UNAFold.pl checks for sir_graph after sir_graph_ng
			--constraints works in UNAFold.pl
			UNAFold.pl prefers PNG to JPEG to GIF for boxplot_ng

3.7a	2010-03-12	salt correction for 1x1 loops corrected

3.8	2010-05-08	several bugs involving circular molecules fixed
			warnings for unsupported long options
			errors for corrupt energy or input files
