The Evidence Layer for Scientific AI

 Quick Analysis Plan Copied

 Long Analysis Plan

Experimental Design: Filopodial EV Transport During Zebrafish Fin Regeneration

Goal. Test whether extracellular vesicles (EVs) are actively transported along filopodia in the regenerating caudal fin blastema, and establish which molecular machinery (Myo10, actin, phosphoinositide lipids) is causal rather than correlational.

1. Live tracking arm

• Amputate adult AB zebrafish caudal fins; image at 12 hpa–3 dpa, coinciding with peak regeneration phosphoproteomic activity (e.g., UCKL1 peaks at 2 dpa; DDRGK1 sustained through 3 dpa) Fin regeneration phosphoproteome time course.
• Label EVs (e.g., CD63/palmGFP reporter) and filopodia (LifeAct/Myo10-Halo); light-sheet or HILO imaging at 1–2 s frame intervals for filopodial transport, mirroring the framework validated in Xenopus RGC filopodia HILO imaging + Granger causality pipeline.
• Quantify: EV velocities (anterograde/retrograde), run lengths, pauses, tip-capture events, filopodial extension rates. Myo10 is the prime candidate motor — its intrinsically disordered cargo-binding regions mediate tunable, multivalent cargo transport and cargo prioritization along filopodia Myo10 IDR-mediated cargo transport.

2. Perturbation controls (split the hypothesis space)

• Motor loss: Myo10 knockdown/CRISPR or dominant-negative M4F domain. Prediction: EV transport abolished with filopodia still present (or Myo10-dependent filopodia absent).
• Actin: Latrunculin B (depolymerization) vs jasplakinolide (stabilization; validated in fin blastema where Jasp drives Yap nuclear translocation within 2 h) Jasp/F-actin perturbation in fin blastema. Prediction: LatB drains tip PI(3,4)P2 and stalls EV movement before gross filopodial collapse.
• Lipid: alpelisib (10 µM, class I PI3K), PITCOIN4 (20 µM, class II PI3K), YU142670 (50 µM, OCRL/INPP5B) at acute doses from the Xenopus study; test non-additivity of combined PI3K+OCRL inhibition as a network control PI3K/OCRL inhibitor dose panel.
• Regeneration-context controls: CA-yap/DN-yap heat-shock transgenics and fgf20a/Sdf1a pathway perturbation (SU5402) to test whether EV transport flux tracks blastema proliferation state rather than filopodia per se FGF-SDF1-CXCR7 feedback in fin regeneration.

3. Bioinformatics/coding plan

• Kymograph extraction and filopodia segmentation (Python: scikit-image + TrackPy; Filopodyan-style pipeline).
• Granger causality between EV position and filopodial tip extension per filopodium, with Benjamini-Hochberg correction — distinguishes motor-driven transport from passive ride on extending filopodia.
• Statistical modeling: mixed-effects models (statsmodels) with fish/filopodium as random effects; n ≥ 10 fish per condition, ≥ 5 filopodia per blastema, pre-registered exclusion criteria.
• Integrate phosphoproteomic time-course markers (UCKL1, DDRGK1) as regeneration-stage covariates to disentangle stage effects from perturbation effects.

Key control gap: a phototoxicity/bleaching control and an inert-nanoparticle (non-EV) tracking control are needed to distinguish active motor transport from passive diffusion or flow — neither supplied record includes such a control. No directly reported zebrafish filopodial EV velocity data exist in the supplied evidence; expected ranges must be established empirically in pilot experiments before power analysis.