Dear Friend, the topic is quite complex especially for someone doing it for the first time. Please have a look at my article from 2011 "Determination of EC50 toxicity data of selected heavy metals toward Heterocypris incongruens... " and recent on "Environmentally oriented models...". I trust You mean EC50 not ED50 (although the latter can also be calculated). Very very very briefly: You need to set up Your experiment depending on the organism used as a model (exposure time, concentration, form of anion, static/semi-static etc.) and expose the organisms (at least in three replicate groups for each concentration level) to series of dilutions of Your toxic metal. Certainly You will run at first blind tests - so called range finding tests, to see approximate concentration levels causing LC50 or change of observed parameter. Then You set up an experiment with increasing concentrations of Your toxicant but in pre-defined concentrations levels and pointing the response (mortality) of observed groups to varying content of toxicants (in mol/dm3 for solutions). The response rarely is linear so either You make logaritmic transformations on the data or make it second order curve. For log of 50% You recalculate the concentration of toxicant that is causing death (change of observed parameter) by 50% (LC50 and EC50 respectively). This is the very very very basics. If I can be of assisstance please let me know. In YouTube You can find channel with mathematical explanations on that but it's very basic.
LC50 is the median lethal dose of a toxic substance, i.e., that dose of a chemical which kills half the members of a tested population. Usually we used Probit Analysis to caculated the LC50.
You can find the detail method of Probit Analysis in the follow attached.
First, you should prepare a geometric serial concentrations of your toxicant, this is the first and most important step in acute toxicity test. Without geometric series you will not be able to use probit regression analysis. I have attached OECD guideline for acute toxicity test (No 203). Please follow each criterium.
When you record your mortality rate at each 24 h, you can apply Probit under DOS or use a statistical software, like SPSS as follow:
Data inputting:
a column for CONCENTRATIONS
2nd column for TOTAL NO OF FISH AT EACH ARENA
3nd, 4th and etc for MORTALITY, e.g. 24h, 48h, 72h and 96h
Data analysis:
Analyze menu >Regression> Probit...
Fill the following boxes:
response frequency= mortality
total observed= Total no of fish...
covariate= Concentrations
Transform= Log base 10
CLICK OK. You will have an output, containing as many details as you like!
Two of my former supervisors developed at package for dose-response-modeling in the statistical free-ware R. It can fit both quantal and continues data depending on experimental design and endpoints and you can easily retrieve ECx-values like EC1, EC5, EC50, EC95 or whatever you need.
One benefit is that you do not need to convert quantal data to a probit scale. Another benefit is that for continues data you can easily use and test several dose-repsonse-models with different numbers of parameters and statistically test them against each other getting p-value for the different models and parameters. Attached is a plot containing three different quantal dose-response curves for different treatments as well as their original article.
Hope this is helpful, feel free to contact me for more advice or help :)
Sahib, if its only the calculation that you are interested and not the experimental design - as the question pertains only to calculation -than for LD50 calculation itself you can use the free EPA toxicity estimation software - http://www.epa.gov/nrmrl/std/qsar/qsar.html, or you can use other statistical software. MiniTab statistical software has a very nice LC50 build in tool with nice graphs. StatFi is another software option with good LD50 option, or alternatively you can calculate LD50 by hand if you know how to work with probit functions.
When you are talking 'calculating' do you mean based on experimental data (which is pretty nice described in the above answers) or do you mean by e.g. qsar approaches, which unfortunately is not that well suited for metal ions. For organics substances the qsar approaches are well-described in the literature, which I a sure you are aware of
your question does not define the endpoint and indicator you are investigating. This is essential as LC50 and ED50 as well as EC50 are quite different between single organism and a community, a death and a symptom of a living organism.
You need to specify your experimental setup before your Q can be answered correctly
Dear dr. Sahib Mohammad bakir the reply of dr Błażej Kudłak is quite sufficient, you would get better answer if you were able to reveal the animal against you are going to test and the metal under consideration. For you calculations either you employee probit analysis or in the preliminary trials, you can use log probit graph paper to avoid computation of data several time, however, at the end you would be required to do probit analysis.
First you have to find range finding test for our organism . Then split the no intervals to eight/ten depending upon you ranges then follow the mortality test . At last use probit tables to plot the graph between log concentration and probit values of mortality to find the lethal concentrations.