# A question about to rerun a .trr file with a .xtc file

**URL:** <https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742>\
**Category:** User discussions\
**Tags:** mdrun\
**Created:** [February 25, 2022, 11:16pm UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742 "2022-02-25T23:16:08Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![liuwq](https://avatars.discourse-cdn.com/v4/letter/l/c37758/32.png) [@liuwq](https://gromacs.bioexcel.eu/u/liuwq)\
**Post date:** [February 25, 2022, 11:16pm UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742/1 "2022-02-25T23:16:08Z")

</div>

GROMACS version: 2018.8  
I used `.xtc` file format to store the trajectories of my previous simulations. But now I need the forces and velocities to do some analysis. The simple way is to start the simulation from the very beginning with a new `.mdp` file, but it is also very time-consuming. I am wondering can I use the existing `.xtc` and the `mdrun -rerun` function to get the `.trr` file. Are the forces and velocities as accurate as the result got by completely running the simulation again?  
Any help would be much appreciated!

---

<div class="post-metadata">

**Author:** ![jalemkul](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/jalemkul/32/18_2.png) [@jalemkul](https://gromacs.bioexcel.eu/u/jalemkul)\
**Post date:** [March 3, 2022, 1:28pm UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742/2 "2022-03-03T13:28:09Z")

</div>

The forces will be accurate. Velocities from reruns are not trustworthy, and I’m not sure if they are even calculated.

---

<div class="post-metadata">

**Author:** ![liuwq](https://avatars.discourse-cdn.com/v4/letter/l/c37758/32.png) [@liuwq](https://gromacs.bioexcel.eu/u/liuwq)\
**Post date:** [March 3, 2022, 11:11pm UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742/3 "2022-03-03T23:11:54Z")

</div>

Thank you very much!  
Then I need to start the simulation from the beginning but set the output format to `.trr`.

---

<div class="post-metadata">

**Author:** ![jalemkul](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/jalemkul/32/18_2.png) [@jalemkul](https://gromacs.bioexcel.eu/u/jalemkul)\
**Post date:** [March 4, 2022, 2:45am UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742/4 "2022-03-04T02:45:19Z")

</div>

You also need to specify a non-zero value of `nstvout` in the `.mdp` file; output options related to bulky `.trr` files are disabled by default, even if you tell `mdrun` to write one, it will only contain the data you request.

---

<div class="post-metadata">

**Author:** ![liuwq](https://avatars.discourse-cdn.com/v4/letter/l/c37758/32.png) [@liuwq](https://gromacs.bioexcel.eu/u/liuwq)\
**Post date:** [March 13, 2022, 4:51pm UTC](https://gromacs.bioexcel.eu/t/a-question-about-to-rerun-a-trr-file-with-a-xtc-file/3742/5 "2022-03-13T16:51:26Z")

</div>

Hello,  
Thank you very much for reminding me!  
I have found that it has separate options for different properties by `nstxout`, `nstvout` and `nstfout`.