Release v1.5.0

A new Simulate tool for planning titrations before the bench, richer exploration of your fit results, batch import of replicates, plain-language messages, and correct handling of units.

New Features

• Simulate tool — preview a titration curve before running it: pick an assay, set values, see the predicted curve live (under Simulate). It lets you:

  • drag concentrations, binding constants, and signal values on live sliders with custom ranges
  • set the titration by a maximum concentration or your own points
  • add optional noise to mimic real scatter
  • overlay a real measurement to compare against the prediction
  • read out host, dye, and complex amounts at every point, following your cursor
  • save setups and export the curve or its data

• Explore your fit results — inspect the whole spread of good fits, not just one number:

  • the table shows a typical value and how widely fits vary
  • switch how the ± is summarized without re-running
  • see every fit's quality and click one to report it instead

• Batch import of replicates — load several files at once as replicates in one session, labelled by file name

• Editable session tabs — tabs are named after the loaded file; double-click to rename

Improvements

• More trustworthy results — reports one of your actual best fits instead of an average that matched no real fit, and keeps strong fits that were previously discarded
• Grouped assay menu — pick a binding-mechanism category, then the variant
• Plain-language messages — errors say what happened and what to do next
• Interface polish — cleaner tabs, a sidebar that no longer gets cut off, and mouse-wheel scrolling across many open tabs

Fixes

• Files read in their stated unit — a file's declared unit is honored; a missing or unknown unit is refused with a clear message instead of producing wrong binding constants
• Coarser units keep your value — switching to a larger unit (µM → M) no longer rounds it to zero
• Clearer import warnings — mismatched concentration points or unequal column lengths are explained, not failed quietly
• Units survive save and reload — fit-result units no longer come back blank
• Sturdier results import — a damaged saved-results file shows a clear message instead of crashing

---

Downloads: macOS, Linux, and Windows builds are attached below.
Full changelog: v1.4.0...v1.5.0

Release v1.4.0

SupraSimFit v1.4.0

This release adds fitting for host–guest systems that bind in two steps, and makes fitted curves render cleanly.

New Features

• Stepwise 1:2 binding (HG2) — fit a host that binds two guests in succession (H + G ⇌ HG, HG + G ⇌ HG₂), recovering both stepwise association constants from a single titration.
• Stepwise 2:1 binding (H2G) — the mirror case, where two hosts bind one guest (H + G ⇌ HG, HG + H ⇌ H₂G).
• Selectable from the assay menu — both appear alongside the existing assays, each with its own host-concentration input and built-in help describing the model and what it fits.

Improvements

• Smooth fit curves — the median-fit line now draws and exports as a smooth curve instead of straight segments between data points.

Release v1.3.0

SupraSimFit 1.3.0

This release adds support for more instrument formats and a built-in update notifier.

New Features

• JASCO import — load JASCO Spectra Manager titration files (.csv) directly.
• EnSight import — load PerkinElmer EnSight plate-reader exports (.csv); multi-channel files get a Channel dropdown to plot and fit each optical channel without re-importing.
• Update notifier — the app tells you when a new version is available and links you straight to the download (you install it yourself).

Improvements

• Plate-reader imports no longer interrupt with a popup — enter concentrations inline; a fit is still blocked until you do.

Release v1.2

SupraSimFit v1.2

The app is now called SupraSimFit. This release is about getting your results out of the app cleanly and making the defaults work for you.

New Features

• One-click Export… (Ctrl+Shift+E) — save raw data, fit results, plots, and your style template in one step.
• Distributions plot export with a layout picker (rows × columns, size, DPI) and live preview.
• Editable axis names in Plot Style, with HTML support; units stay auto-managed.
• Edit concentrations directly in the Data panel — the pop-up dialog is gone, edits apply immediately.
• Friendlier data loading: European CSVs, fuzzy column headers, and headerless files just work.

Improvements

• Fit-results overlay (Ka, R², parameters) now shows by default after a fit.
• Per-replica fitting is the new default, giving honest cross-replica uncertainty out of the box.

Fixes

• DBA dye-to-host signal corrected: dye-into-host titrations previously used the wrong species concentration and gave the wrong curve. Host-to-dye fits are unaffected.
• High-resolution plot exports now render the legend and parameter annotations at the right size.
• Windows app icon is now shown correctly on the executable and in the taskbar.
• Stray scroll-wheel edits no longer change numeric input fields when you accidentally scroll over them.

Release v1.1

style(gui): refine QGroupBox title styling

Adjust title top offset, padding, and background color to align with
InfoGroupBox button placement and window palette.

Release v1.0.1

v1.0.1 — data-driven parameter rescaling

Release v1.0

fix(build): rewrite PyInstaller spec and release workflow for PyQt6

Add run_app.py as the bundled entry point and update the spec to
declare the pyqtgraph, scipy, and pint hidden imports, bundle the
demo datasets, and ship a proper macOS Info.plist. Drive the CI
build via astral-sh/setup-uv + uv run --with pyinstaller, and use
ditto to package the .app so Qt framework symlinks survive.

Release v0.4.2

Update legend positioning in plots to avoid overlap with the data points

Release v0.4.1

Remove CellWriter import from main_interface.py

Release v0.4

Add preprocessing section with plot replica and BMG to TXT converter …