#!/usr/bin/env python3 """ Generate Sepia Fig. 1 / DNA template figure from real processed CSV data. Inputs expected in ./data/ or in the current working directory: - 260514_fig2_endpoint_well_table.csv - 260514_fig2_processed_long.csv The figure uses real replicate-level endpoint data and kinetic replicate data. Endpoint means and SD error bars are recomputed from the endpoint well table; kinetic traces are recomputed from the long-format replicate table. """ from __future__ import annotations from pathlib import Path import math import numpy as np import pandas as pd import matplotlib matplotlib.use("Agg") import matplotlib.pyplot as plt from sepia_figure_style import ( LINEAR_GREEN, PLASMID_GREEN, TEXT, style_axis, ) # Match the current Shopify product-page Fig. 1 image slot. # The OpenCFPS template declares width="1100" height="432" for this figure. SCRIPT_VERSION = "2026-06-09.1" FIG_W_PT = 1100.0 FIG_H_PT = 432.0 FIG_W_IN = FIG_W_PT / 72.0 FIG_H_IN = FIG_H_PT / 72.0 # Manual layout tuned for the wide, shallow Shopify figure slot. LEFT_AX = [0.075, 0.205, 0.365, 0.620] RIGHT_AX = [0.548, 0.205, 0.355, 0.620] TARGETS = { ("Plasmid", 4.0): {"label": "Plasmid", "color": PLASMID_GREEN, "marker": "o"}, ("Linear", 8.0): {"label": "Linear", "color": LINEAR_GREEN, "marker": "s"}, } LABEL_BOX = None def find_input_dir() -> Path: """Find input CSVs next to the script or in ./data.""" here = Path(__file__).resolve().parent candidates = [here / "data", here, Path.cwd() / "data", Path.cwd()] for candidate in candidates: if (candidate / "260514_fig2_endpoint_well_table.csv").exists() and ( candidate / "260514_fig2_processed_long.csv" ).exists(): return candidate raise FileNotFoundError( "Could not find required CSVs. Put the two input CSV files in ./data or next to this script." ) def summarize_endpoint(endpoint_wells: pd.DataFrame) -> pd.DataFrame: required = {"DNA_Type", "DNA_Actual_nM", "RFE_uM"} missing = required - set(endpoint_wells.columns) if missing: raise ValueError(f"Endpoint table missing columns: {sorted(missing)}") summary = ( endpoint_wells.groupby(["DNA_Type", "DNA_Actual_nM"], as_index=False) .agg(N=("RFE_uM", "count"), Mean_RFE_uM=("RFE_uM", "mean"), SD_RFE_uM=("RFE_uM", "std")) .sort_values(["DNA_Type", "DNA_Actual_nM"]) ) summary["SD_RFE_uM"] = summary["SD_RFE_uM"].fillna(0.0) summary["x_nM"] = summary["DNA_Actual_nM"].astype(float) return summary def summarize_kinetics(long_df: pd.DataFrame) -> pd.DataFrame: required = {"Assay", "Time_h", "DNA_Type", "DNA_Actual_nM", "RFE_uM", "Raw_610_650_RFU"} missing = required - set(long_df.columns) if missing: raise ValueError(f"Long kinetic table missing columns: {sorted(missing)}") df = long_df.loc[long_df["Assay"].eq("TXTL")].copy() keep = pd.Series(False, index=df.index) for dna_type, dna_nM in TARGETS: keep |= df["DNA_Type"].eq(dna_type) & np.isclose(df["DNA_Actual_nM"], dna_nM) df = df.loc[keep] if df.empty: raise ValueError("No kinetic rows found for target plasmid 4 nM and linear 8 nM conditions.") kinetic = ( df.groupby(["Time_h", "DNA_Type", "DNA_Actual_nM"], as_index=False) .agg( N=("RFE_uM", "count"), Mean_RFE_uM=("RFE_uM", "mean"), SD_RFE_uM=("RFE_uM", "std"), Mean_MgAptamer_RFU=("Raw_610_650_RFU", "mean"), SD_MgAptamer_RFU=("Raw_610_650_RFU", "std"), ) .sort_values(["DNA_Type", "DNA_Actual_nM", "Time_h"]) ) return kinetic.fillna(0.0) def make_figure(input_dir: Path, out_svg: Path, out_png: Path | None = None) -> None: endpoint_wells = pd.read_csv(input_dir / "260514_fig2_endpoint_well_table.csv") long_df = pd.read_csv(input_dir / "260514_fig2_processed_long.csv") endpoint = summarize_endpoint(endpoint_wells) kinetic = summarize_kinetics(long_df) # Write computed summaries next to the figure so the plotted values can be inspected. summary_dir = out_svg.parent summary_dir.mkdir(parents=True, exist_ok=True) endpoint.to_csv(summary_dir / "sepia_fig1_endpoint_summary_used.csv", index=False) kinetic.to_csv(summary_dir / "sepia_fig1_kinetics_summary_used.csv", index=False) fig = plt.figure(figsize=(FIG_W_IN, FIG_H_IN), facecolor="none") # ------------------------------------------------------------------ # Panel A: endpoint DNA concentration response, linear x axis. # ------------------------------------------------------------------ ax_a = fig.add_axes(LEFT_AX) style_axis(ax_a) order = ["Plasmid", "Linear"] for dna_type in order: sub = endpoint.loc[endpoint["DNA_Type"].eq(dna_type)].sort_values("x_nM") color = PLASMID_GREEN if dna_type == "Plasmid" else LINEAR_GREEN marker = "o" if dna_type == "Plasmid" else "s" linestyle = "-" face = color if dna_type == "Plasmid" else "white" ax_a.errorbar( sub["x_nM"], sub["Mean_RFE_uM"], yerr=sub["SD_RFE_uM"], color=color, linestyle=linestyle, linewidth=3.4, marker=marker, markersize=9.5, markerfacecolor=face, markeredgecolor=color, markeredgewidth=1.55, capsize=4.5, elinewidth=1.8, capthick=1.8, zorder=3, ) ax_a.set_xticks([0, 4, 8, 16, 32]) ax_a.set_xlim(-0.8, 33.5) ax_a.set_ylim(0, 50) ax_a.set_yticks([0, 25, 50]) ax_a.set_xlabel("DNA concentration (nM)", labelpad=8) ax_a.set_ylabel("Relative Fluorescein Equivalents (µM)", labelpad=8) ax_a.text( 7.2, 44.8, "Plasmid DNA", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) ax_a.text( 19.2, 27.0, "Linear DNA", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) # ------------------------------------------------------------------ # Panel B: kinetic traces, two selected template conditions. # TL = GFP/RFE on left axis; TX = MG aptamer fluorescence on right axis. # ------------------------------------------------------------------ ax_b = fig.add_axes(RIGHT_AX) style_axis(ax_b) ax_b2 = ax_b.twinx() style_axis(ax_b2, show_right_spine=True) ax_b2.grid(False) ax_b2.tick_params(axis="y", color=TEXT, labelcolor=TEXT, length=5.5, width=1.35) for (dna_type, dna_nM), spec in TARGETS.items(): sub = kinetic.loc[ kinetic["DNA_Type"].eq(dna_type) & np.isclose(kinetic["DNA_Actual_nM"], dna_nM) ].sort_values("Time_h") t = sub["Time_h"].to_numpy(float) tl = sub["Mean_RFE_uM"].to_numpy(float) tl_sd = sub["SD_RFE_uM"].to_numpy(float) tx = sub["Mean_MgAptamer_RFU"].to_numpy(float) tx_sd = sub["SD_MgAptamer_RFU"].to_numpy(float) color = spec["color"] marker = spec["marker"] marker_face = color if dna_type == "Plasmid" else "white" # TL / translation / GFP output. ax_b.plot( t, tl, color=color, linestyle="-", linewidth=3.4, marker=marker, markersize=6.2, markevery=3, markerfacecolor=marker_face, markeredgecolor=color, markeredgewidth=1.35, zorder=4, ) # Subtle SD bands show replicate variation without extra chart furniture. tl_band_color = color ax_b.fill_between(t, tl - tl_sd, tl + tl_sd, color=tl_band_color, alpha=0.10, linewidth=0, zorder=2) # TX / transcription / MG aptamer fluorescence output. tx_color = color ax_b2.plot( t, tx, color=tx_color, linestyle=(0, (1.2, 2.2)), linewidth=2.85, marker=marker, markersize=5.2, markevery=3, markerfacecolor=marker_face, markeredgecolor=tx_color, markeredgewidth=1.2, zorder=3, ) ax_b2.fill_between(t, tx - tx_sd, tx + tx_sd, color=tx_color, alpha=0.10, linewidth=0, zorder=1) x_max = float(kinetic["Time_h"].max()) ax_b.set_xlim(0, math.ceil(x_max)) ax_b.set_xticks([0, 6, 12]) ax_b.set_ylim(0, 50) ax_b.set_yticks([0, 25, 50]) ax_b2.set_ylim(0, 15000) ax_b2.set_yticks([0, 7500, 15000]) ax_b2.set_yticklabels([f"{int(v):,}" for v in [0, 7500, 15000]]) ax_b.set_xlabel("Time (h)", labelpad=8) ax_b.set_ylabel("Relative Fluorescein Equivalents (µM)", labelpad=8) ax_b2.set_ylabel("MG aptamer fluorescence (RFU)", labelpad=8) ax_b.text( 7.25, 52.2, "Plasmid TL", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) ax_b.text( 8.35, 27.2, "Linear TL", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) ax_b2.text( 0.70, 12750, "Plasmid TX", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) ax_b2.text( 0.58, 9200, "Linear TX", color=TEXT, fontsize=17, fontweight="bold", va="center", bbox=LABEL_BOX, ) for axis in (ax_a, ax_b, ax_b2): axis.xaxis.label.set_fontweight("bold") axis.yaxis.label.set_fontweight("bold") for tick_label in axis.get_xticklabels() + axis.get_yticklabels(): tick_label.set_fontweight("bold") out_svg.parent.mkdir(parents=True, exist_ok=True) fig.savefig(out_svg, format="svg", transparent=True, bbox_inches=None, metadata={"Date": None}) if out_png is not None: fig.savefig(out_png, dpi=220, transparent=True, bbox_inches=None) manifest = { "figure": out_svg.name, "script": Path(__file__).name, "script_version": SCRIPT_VERSION, "width_pt": FIG_W_PT, "height_pt": FIG_H_PT, "site_slot": "OpenCFPS product template Fig. 1 asset; template image width=1100 height=432", "endpoint_source": "260514_fig2_endpoint_well_table.csv", "kinetic_source": "260514_fig2_processed_long.csv", "endpoint_summary": "sepia_fig1_endpoint_summary_used.csv", "kinetics_summary": "sepia_fig1_kinetics_summary_used.csv", "endpoint_error_bars": "sample SD across replicate wells", "kinetic_traces": "mean across replicate wells; shaded bands are sample SD for TL and TX traces", "kinetic_conditions": {"plasmid_nM": 4.0, "linear_nM": 8.0}, "endpoint_x_axis": "linear DNA concentration in nM", "right_y_axis": "MG aptamer fluorescence (RFU), computed from Raw_610_650_RFU", "visual_encoding": "DNA type controls trace color and marker in both panels: plasmid is dark green with filled circle markers, linear is light green with open square markers; all direct labels are black; TX traces use dotted lines.", "theme_notes": "Sepia site style: Inter/Lexend-like sans-serif, transparent background, black label text, coordinated green DNA-type palette." } import json (summary_dir / "sepia_fig1_reproducibility_manifest.json").write_text(json.dumps(manifest, indent=2)) plt.close(fig) def main() -> None: input_dir = find_input_dir() out_dir = Path(__file__).resolve().parent / "outputs" make_figure( input_dir=input_dir, out_svg=out_dir / "sepia_fig1_dna_endpoints_kinetics.svg", out_png=out_dir / "sepia_fig1_dna_endpoints_kinetics.png", ) print(f"Wrote {out_dir / 'sepia_fig1_dna_endpoints_kinetics.svg'}") if __name__ == "__main__": main()