Congrats Rita, and thanks for everything you did at Nature Methods! Hopefully our paths cross again at some point 😀

Openly accessible here: rdcu.be/dv1sr !

Finally, huge kudo's to all my labs in all three countries (WUR, NL, CarnegieMellon, USA, and UniBonn GER), and to funding from the #Humboldt Postdoc Fellowship @humboldt-foundation.de , Bonn Argelander program, VLAG@WUR, NSF, CarnegieMellon, and UniBonn! (8/8)

This method wouldn't be here without @uendesfelder.bsky.social , @hohlbeinlab.bsky.social and Dr. Turkowyd, and I love the improvements during the review process led by @ritastrack.bsky.social - it increased the fundamental, mathematical underpinning of TARDIS ánd allows more flexibility in fitting!

TARDIS promises to open up the possibility of performing spt data analysis in/with wildly novel conditions/probes. It will also directly benefit from all future endeavors in (mobile) particle localization at high density. Try TARDIS yourself here: github.com/kjamartens/T...

We used TARDIS to look at the in vivo movement of RNA polymerase, and show that we can easily lower the measurement time by a factor of ~5. This was limited by the required high-density localization, NOT by TARDIS performance. (5/8)

We went to crazy lengths to try to overwhelm TARDIS, but only managed to get a 'suboptimal' error of ~5% once we (far) surpassed the limits of mobile single-molecule localization AND had only 1.5 loc/traj on average AND > 50% of the dataset was pure noise (!) (4/8)

This concept accurately obtains either an analytically fitted distribution (anything is possible here), or a jump-distance histogram. It is far more robust than any tracking algorithm (at high complexity), and constantly surprised us with its robustness. (3/8)

TARDIS is a conceptually new method to perform single-particle tracking analysis: all localizations are compared to themselves with a time-shift. The intraparticle population is separated from the interparticle distribution by observing time delays longer than track lengths: (2/8)