BMDx: dose response analysis for expression data

1. Input gene lists

1) Organisms

human

mouse

rat

2) GeneID

symbols

ensemble

entrez

2. Functional annotation parameters

3. Display parameters

Aggregation Function

min

max

mean

median

Choose Values Type

Pvalue

GenesModifications

GenesModifications_PValue

Plot modification

value

sign

4. Filter BMD Values

5. Download gene lists as excel file

Download

Response Level

Assumption of Constant Variance

Lack-of-fit PValue Th:

Lowest dose filter:

Highest dose filter:

Number of cores:

Select the BMD analysis setting

Models Description

Linear Model

The formula for the linear model is $$ f(dose) = \beta_0 + \beta_1 dose $$ The linear model is a special case of the polynomial model, with $$n=1$$

Polynomial Model (Quadratic/Cubic)

The formula for the polynomial model is $$ f(dose) = \beta_0 + \beta_1 dose + \beta_2 dose^2 + \ldots + \beta_n dose^n $$ Here n is the degree of the polynomial. The user can choose between $$n = 2, 3$$

Power Model

The formula for the power model is $$ f(dose) = \beta_0 + (dose)^\delta $$ The user can choose between $$\delta = 2, 3, 4$$

Exponential Model

The formula for the exponential model is $$ f(dose) = \beta_0 + exp(dose) $$

Hill Model

The formula for the hill model is $$ f(dose) = \beta_0 + \dfrac{dose^n}{Kd + dose^n} $$ The user can choose between $$n = 0.5,1,2,3,4,5$$ while Kd is fixed to 10.

Asymptotic Regression

The formula for the asymptotic regression model is the following: $$f(dose) = c + (d-c) \times (1-exp(-dose/e)) $$ The parameter c is the lower limit (at x=0), the parameter d is the upper limit and the parameter e>0 is determining the steepness of the increase of dose. The AR.3 model is the one depending from c, d and e parameters. The AR.2 model depends only on d and e parameters, while c is set to zero

Michaelis-Menten Model

The model is defined by the three-parameter model (MM.3) function $$f(dose, (c, d, e)) = c + \dfrac{d-c}{1+(e/dose)}$$ It is increasing as a function of the dose, attaining the lower limit c at dose 0 (x=0) and the upper limit d for infinitely large doses. The parameter e corresponds to the dose yielding a response halfway between c and d. The common two-parameter Michaelis-Menten model (MM.2) is obtained by setting c equal to 0.

BMD Values

Lack of fit Pvalues

BMD/BMDL

Fitted models

Gene by Time Point

Venn diagram responsive genes

1. Data Selection

2. Plot section

3. Download Selection

Use scrollbars to navigate and see the whole map

Enrichment

1. Input gene lists

2. Functional annotation parameters

3. Display parameters

4. Filter BMD Values

5. Download gene lists as excel file

Compute BMD Value

Models Description

Filter Genes by Anova

Filter Genes by Trend Test

Import Gene Expression Table

Import Phenotype Data

1. Data Selection

2. Plot section

3. Download Selection

Use scrollbars to navigate and see the whole map