Hello,
I have observed that the more templates I include in the homology modeling the least Ca-Ca distance restraints I obtain. I can't understand this since I would expect to obtain restraints from missing residues, homology, etc from more templates and therefore more restraints imposed into my model. What could an explanation to this?
Thank you!
Best,
Mariana Gonzalez Medina
On 4/16/20 8:41 AM, Mariana GONZALEZ MEDINA wrote: > I have observed that the more templates I include in the homology > modeling the least Ca-Ca distance restraints I obtain. I can't > understand this since I would expect to obtain restraints from missing > residues, homology, etc from more templates and therefore more > restraints imposed into my model. What could an explanation to this?
A given CA-CA distance will only have a single restraint on it, regardless of how many templates you have (although it will become multimodal). (If Modeller were simply to add restraints for each template you would end up with a model that is an average of all templates, which is not physically correct - if one template says a CA-CA distance is 10A and another says it is 20A, the distance in the model should be either 10A or 20A, not 15A.)
Ben Webb, Modeller Caretaker
Hi,
I understand now, I was only taking into account Ca-Ca distances for which the mathematical from was a Gaussian but if I use multiple templates the restraint becomes multimodal and the from is a spline. If I were to plot the distances with a gaussian from I would be plotting the probability density and the distance values. What are the values that I get for the spline?
Thank you.
Best,
Mariana
________________________________ From: Modeller Caretaker modeller-care@salilab.org Sent: Thursday, April 16, 2020 7:09:15 PM To: Mariana GONZALEZ MEDINA; modeller_usage@salilab.org Subject: Re: [modeller_usage] Ca-Ca restraints
On 4/16/20 8:41 AM, Mariana GONZALEZ MEDINA wrote: > I have observed that the more templates I include in the homology > modeling the least Ca-Ca distance restraints I obtain. I can't > understand this since I would expect to obtain restraints from missing > residues, homology, etc from more templates and therefore more > restraints imposed into my model. What could an explanation to this?
A given CA-CA distance will only have a single restraint on it, regardless of how many templates you have (although it will become multimodal). (If Modeller were simply to add restraints for each template you would end up with a model that is an average of all templates, which is not physically correct - if one template says a CA-CA distance is 10A and another says it is 20A, the distance in the model should be either 10A or 20A, not 15A.)
Ben Webb, Modeller Caretaker -- modeller-care@salilab.org https://salilab.org/modeller/ Modeller mail list: https://salilab.org/mailman/listinfo/modeller_usage
On 4/20/20 1:31 AM, Mariana GONZALEZ MEDINA wrote: > I understand now, I was only taking into account Ca-Ca distances for > which the mathematical from was a Gaussian but if I use multiple > templates the restraint becomes multimodal and the from is a spline. If > I were to plot the distances with a gaussian from I would be plotting > the probability density and the distance values. What are the values > that I get for the spline?
It is exactly the same for a multimodal restraint - it is a probability density on the distance. A cubic spline is used because it is faster to evaluate than multiple Gaussians.
Ben Webb, Modeller Caretaker
participants (2)
-
Mariana GONZALEZ MEDINA -
Modeller Caretaker