Re: How to fit a Hill model using braidrm package?

varin sacha via R-help <[email protected]>
Newsgroups gmane.comp.lang.r.general
Message-ID <[email protected]> 
Hi,


the drc package provides a Hill function. You could try this : 



library(drc)

# Fit Hill function

hill_mod <- drm(

  Response ~ Dose,

  data = df,

  fct = hill(names = c("Top", "Bottom", "Kd", "nH"))

)

summary(hill_mod)

# Plot (log-scale on x-axis)

plot(hill_mod, log = "x")





The 4 parameters are:

Top : max response at low dose
Bottom : min response at high dose
Kd : inflection point
nH : Hill slope 
Your data has Response ≈ 895 at low Dose and goes down to ≈ 850 at high Dose, so the Hill curve must be decreasing.

That means:

Top ≈ 895 
Bottom ≈ 850 
Kd ≈ midpoint dose (where Response ≈ halfway between 895 and 850 : ~872). Looking at your doses, this is around 10^{-6}.
nH ≈ slope (start with 1, adjust later)


Using your data

library(drc)

# Your data
df = data.frame(
  Response = c(890.72, 890.94, 880.16, 895.46, 890.8, 884.15,
               895.63, 887.22, 879.57, 894.72, 888.91, 878.89,
               895.49, 890.83, 882.27, 893.59, 889.92, 881.59,
               892.53, 891.76, 880.98, 895.06, 890.32, 881.45,
               897.21, 886.35, 876.19, 889.27, 878.11, 868.32,
               876.05, 866.57, 859.16, 850.53, 857.96, 859.34,
               862.02, 859.04, 859.73, 858.36, 863.64, 861.19),
  Dose = c(0.0000000015, 0.0000000015, 0.0000000015, 0.000000003,
           0.000000003, 0.000000003, 0.000000006, 0.000000006,
           0.000000006, 0.000000012, 0.000000012, 0.000000012,
           0.000000024, 0.000000024, 0.000000024, 0.000000048,
           0.000000048, 0.000000048, 0.000000095, 0.000000095,
           0.000000095, 0.00000018, 0.00000018, 0.00000018,
           0.00000038, 0.00000038, 0.00000038, 0.00000078,
           0.00000078, 0.00000078, 0.0000015, 0.0000015,
           0.0000015, 0.000007, 0.000007, 0.000007,
           0.000025, 0.000025, 0.000025, 0.00005, 0.00005, 0.00005)
)

# Fit Hill function with starting values
hill_mod <- drm(
  Response ~ Dose,
  data = df,
  fct = hill(names = c("Top", "Bottom", "Kd", "nH")),
  start = c(Top = 895, Bottom = 850, Kd = 1e-6, nH = 1)
)

summary(hill_mod)

# Plot fit on log-scale x-axis
plot(hill_mod, log = "x", ylim = c(840, 900))



Then, you can extract the fitted parameters (including Kd and nH) directly.



# Coefficients (parameter estimates)

coef(hill_mod)



# More detailed summary with SEs and p-values

summary(hill_mod)



# Confidence intervals for parameters

confint(hill_mod)



# Extract just the estimated Kd (dissociation constant)

Kd_est <- coef(hill_mod)["Kd:(Intercept)"]



# Extract Hill slope (nH)

nH_est <- coef(hill_mod)["nH:(Intercept)"]

Kd_est

nH_est



Best,






> Le 25 sept. 2025 à 13:41, Luigi Marongiu <[email protected]> a écrit :
> 
> Hello,
> it was an assembler problem of the machine. I now managed to install
> the package, I tried to fit a 4 or 5 parameters model but the fitting
> is not good:
> 
> ```
> df = data.frame(Response =    c(890.72,    890.94,    880.16,
> 895.46,    890.8,    884.15,
>                             895.63,    887.22,    879.57,    894.72,
>  888.91,    878.89,
>                             895.49,    890.83,    882.27,    893.59,
>  889.92,    881.59,
>                             892.53,    891.76,    880.98,    895.06,
>  890.32,    881.45,
>                             897.21,    886.35,    876.19,    889.27,
>  878.11,    868.32,
>                             876.05,    866.57,    859.16,    850.53,
>  857.96,    859.34,
>                             862.02,    859.04,    859.73,    858.36,
>  863.64,    861.19),
>                Dose =    c(0.0000000015,    0.0000000015,
> 0.0000000015,    0.000000003,
>                         0.000000003,    0.000000003,    0.000000006,
>  0.000000006,
>                         0.000000006,    0.000000012,    0.000000012,
>  0.000000012,
>                         0.000000024,    0.000000024,    0.000000024,
>  0.000000048,
>                         0.000000048,    0.000000048,    0.000000095,
>  0.000000095,
>                         0.000000095,    0.00000018,    0.00000018,
> 0.00000018,    0.00000038,
>                         0.00000038,    0.00000038,    0.00000078,
> 0.00000078,    0.00000078,
>                         0.0000015,    0.0000015,    0.0000015,
> 0.000007,    0.000007,
>                         0.000007,    0.000025,    0.000025,
> 0.000025,    0.00005,    0.00005,    0.00005)
> )
> plot(Response~log10(Dose), df)
> library(drc)
> mod = drm(Response~Dose, data = df, fct=L.4())
> plot(mod)
> ```
> How can I set a Hill function with `drm`?
> Thank you
> 
>> On Thu, Sep 25, 2025 at 1:01 PM varin sacha <[email protected]> wrote:
>> 
>> Hi,
>> 
>> What about directly calling the model function with basicdrm library ?
>> 
>> library(basicdrm)
>> 
>> # Fit Hill model
>> m <- drm(Response ~ Dose, data = df, fct = hillfct())
>> 
>> summary(m)
>> 
>> # Predict & plot
>> plot(m, log = "x")
>> 
>> Best,
>> SV
>> 
>>>> Le 25 sept. 2025 à 10:12, varin sacha via R-help <[email protected]> a écrit :
>>> 
>>> 
>>> Hi,
>>> Could you paste me the exact error message you’re getting when trying install.packages("drc")
>>> 
>>> 
>>>> Le 25 sept. 2025 à 09:02, Luigi Marongiu <[email protected]> a écrit :
>>>> 
>>>> I got the same errors:
>>>> ```
>>>>> library(basicdrm)
>>>>> findBestHill(Response~Dose, df)
>>>> Error in findBestHill(Response ~ Dose, df) :
>>>> could not find function "findBestHill"
>>>>> evalHillEqn(Response~Dose, df)
>>>> Error in evalHillEqn(Response ~ Dose, df) :
>>>> could not find function "evalHillEqn"
>>>> ```
>>>> 
>>>>>> On Wed, Sep 24, 2025 at 1:47 PM peter dalgaard <[email protected]> wrote:
>>>>> 
>>>>> They're in basicdrm, not braidrm...
>>>>> 
>>>>> -pd
>>>>> 
>>>>>>> On 24 Sep 2025, at 11:52 , Luigi Marongiu <[email protected]> wrote:
>>>>>> 
>>>>>> Hello,
>>>>>> I have a set of data coming from a dissociation experiment
>>>>>> (protein/ligand). Since the data is required to calculate the constant
>>>>>> of dissociation (Kd) of this pair, I am looking for a way of fitting a
>>>>>> Hill function to the data.
>>>>>> I have seen that the package braidrm
>>>>>> (https://cran.r-project.org/web/packages/braidrm/index.html) provides
>>>>>> this function, but when I launch the function `evalHillEqn`,
>>>>>> `findBestHill` and so forth I get the error of function not found.
>>>>>> Yet, the package is given as properly installed by the system.
>>>>>> How can I run this package?
>>>>>> Is there an alternative way to fit a Hill function to these data?
>>>>>> Thank you
>>>>>> 
>>>>>> ```
>>>>>> df = data.frame(Response =    c(890.72,    895.46,    895.63,
>>>>>> 894.72,    895.49,    893.59,
>>>>>>             892.53,    895.06,    897.21,    889.27,    876.05,
>>>>>> 857.96,    862.02,    858.36,
>>>>>>             890.94,    890.8,    887.22,    888.91,    890.83,
>>>>>> 889.92,    891.76,    890.32,
>>>>>>             886.35,    878.11,    866.57,    859.04,    863.64,
>>>>>> 880.16,    884.15,    879.57,
>>>>>>             878.89,    882.27,    881.59,    880.98,    881.45,
>>>>>> 876.19,    868.32,    859.16,
>>>>>>             850.53,    853.21,    859.34,    859.73,    861.19),
>>>>>>             Dose =    c(0.0000000015,    0.000000003,
>>>>>> 0.000000006,    0.000000012,
>>>>>>                    0.000000024,    0.000000048,    0.000000095,
>>>>>> 0.00000018,
>>>>>>                    0.00000038,    0.00000078,    0.0000015,
>>>>>> 0.000013,    0.000025,
>>>>>>                    0.00005,    0.0000000015,    0.000000003,
>>>>>> 0.000000006,
>>>>>>                    0.000000012,    0.000000024,    0.000000048,
>>>>>> 0.000000095,
>>>>>>                    0.00000018,    0.00000038,    0.00000078,
>>>>>> 0.0000015,    0.000025,
>>>>>>                    0.00005,    0.0000000015,    0.000000003,
>>>>>> 0.000000006,
>>>>>>                    0.000000012,    0.000000024,    0.000000048,
>>>>>> 0.000000095,
>>>>>>                    0.00000018,    0.00000038,    0.00000078,
>>>>>> 0.0000015,    0.000003,
>>>>>>                    0.000006,    0.000013,    0.000025,    0.00005)
>>>>>> )
>>>>>> plot(Response~log10(Dose), df)
>>>>>> library(braidrm)
>>>>>> evalHillEqn(Response~Dose, df)
>>>>>> findBestHill(Response~Dose, df)
>>>>>> ```
>>>>>> 
>>>>>> ______________________________________________
>>>>>> [email protected] mailing list -- To UNSUBSCRIBE and more, see
>>>>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>>>>> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
>>>>>> and provide commented, minimal, self-contained, reproducible code.
>>>>> 
>>>>> --
>>>>> Peter Dalgaard, Professor,
>>>>> Center for Statistics, Copenhagen Business School
>>>>> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
>>>>> Phone: (+45)38153501
>>>>> Office: A 4.23
>>>>> Email: [email protected]  Priv: [email protected]
>>>>> 
>>>> 
>>>> 
>>>> --
>>>> Best regards,
>>>> Luigi
>>>> 
>>>> ______________________________________________
>>>> [email protected] mailing list -- To UNSUBSCRIBE and more, see
>>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>>> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
>>>> and provide commented, minimal, self-contained, reproducible code.
>>> 
>>>   [[alternative HTML version deleted]]
>>> 
>>> ______________________________________________
>>> [email protected] mailing list -- To UNSUBSCRIBE and more, see
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide https://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>> 
> 
> 
> --
> Best regards,
> Luigi

	[[alternative HTML version deleted]]
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.