I made myself a software using QBASIC in DOS environment. With this software I can calculate IC50 and confidential limits. But you have to have DOS at 32 bits on your operative system
I usually use the program called Prism to calculate IC50s. But you can also use Excel. The easiest way is to calculate from the graph with the derived equation.
sorry, but excel des not give you a good possibility for IC50 values measurements. Most of correlations/relationships of 'dose-effect' has a non-linear character therefore you should use more specialised statistical programms, such as. STATISTICA or much better GraphPad Prism. Both of these proramms you can downloaded for free per 1 month. If you are interested in using GraphPad Prism, please send me your e-mail, then I will send you a very good book in english concerning linear and non-linear regression as well as IC50 values.
You can use a sigmoidal curve to evaluate the LD50 or IC50. Just plot the concentration against the response (scatter plot), then adjust a sigmoidal curve and you will get the LD50 (using a logistic regression model). If you want I can send you the equation. Usually any stat or plot program can provide you this. You can quickly have a look at one possible equation in http://en.wikipedia.org/wiki/EC50
Different relationships between our data demand to be treated in the individual way. I remember that I spent about 2 weeks on reading theory about shapes and equations concerning different IC50 measurements before I was ready to calculate these values for my studies. Thus, I suggest to better know the character of data and after that to try calculate IC50 values. It is not so trivial and so easy thing. :)
You can use R CRAN software (GPL, free) with DRC package (Dose-response curve : http://cran.r-project.org/web/packages/drc/drc.pdf) to perform non linear regression and calculate ED50 in a semi-automatic manner. It's not as difficult as it appear at first glance.
Have a look on R, it's a free, GPL and very powerfull stat software.
Read my paper whose full text is available herein. Rath S, Sahu MC, Dubey D, NK Debata, RN Padhy (2011). Which values should be used as the lethal concentration 50 (LC50) with bacteria? Inter-Disciplin. Sci.- Comput. Life Sci. 3:138-143. (Springer).
one important issue that was mentioned here was the pattern of data. This is the first thing to be aware of, becasue the aim is to fit the best regression possible. If is is a line, is a line, if it is a curve there are several curves available.
you can use excel, but the stat packages mentioned above are really great at giving you what you are looking for. You will have to log transform your concentrations of whatever it is you are using and place this on your x axis, and normalize your data into say % activity on the y-axis. That is to say transform raw data so that you have a range from 0 to 100. plot this using excel and I believe you can deteremine what the LD50,IC50 , and LC50 are. Just make sure you know the difference as some of these have different meanings are are calculated a bit differently. I hope this has helped. I perfer to use GraphPad as it is much better at determining the values.
You can use Excel or SPSS easily if the relationship of 'dose-effect' has a linear pattern, otherwise specialized statistical programs, such as: GraphPad Prism or STATISTICA would be better.
First of all, you need to describe the response data. Is it binary, dead/alive, or is it quantitative, like an Optical Density (OD), Counts per Minute (CPM), or similar. Second, you need to carefully define what a 0% response is and what a 100% response is. That is the only reliable way to know what a true 50% response is. For cell death that could be none died or all died. But for the quantitative variables, you have to have both a control for 0% response (that is usually no drug but everything else the same) and a control for 100% that describes the maximal possible response. Then you can think about the calculations. For quantitative data, you can use the Solver in Excel to do the calculations if that is your only tool. If your data is quantitative and your data includes values for 0% and 100% controls, you could send it to me ([email protected]) and I could send you an Excel workbook with the calculations.
The technically sound method as statistically accepted is the 4 Parameter Logisitic Equation (Sigmoidal) fit to find IC 50. While statistical packages can do this - and cost a bit - you can use at least two FREE sites to calculate EC50.
1. http://www.readerfit.com/
2. http://www.changbioscience.com/stat/ec50.html
Attached is an article on guidelines for EC50/IC50.
Mendes LF, Zambotti-Villela L, Colepicolo P, Marinho-Soriano E, Stevani CV and Yokoya NS (2013) “Metal Cation Toxicity in the Alga Gracilaria domingensis as Evaluated by the Daily Growth Rates in Synthetic Seawater”, Journal of Applied Phycology, DOI 10.1007/s10811-013-0036-1.
Mendes LF and Stevani CV (2010) “Evaluation of Metal Toxicity by a Modified Method Based on the Fungus Bioluminescence in Agar Medium”, Environmental Toxicology and Chemistry, 29, 320-326.
Calculation of LC 50 or LD 50 varies from drug to drug. LC 50 is generally applied for the dose in liquid form(aquatic) while while LD50 is direct dose dependent (solid form). hence the nature and quality of drug varies and reciepent cell behaves accordingly
LC 50 = lethal concentration, where 50% animals are killed by the drug- average of 3-5 expts.
LD 50= lethal concentration of the drug, where you prepare the drug solution based on its molecular weight. and using this molar solution- you determine 50% mortality.
This is the basic information, If you understand this then you can follow all other calculation- by Regression equation.
you have to make a graph by plotting percent inhibition/ percent mortality on Y axis and concentration/ dose on X axis by using excel. Then you can calculate LC50, LD50 or IC50 either manually by using just a ruler or by an equation Y=mX+c. You can get the equation of your result by- place your mouse on curve> press right button> add trend line> options> show equation on chart. Then put Y=50 and calculate the value of X i. e. LD50 and so on.
The way of calculation for LC50, LD50, IC50 values of drugs against cultured cancerous cells is the same usually applied for other cultured cells: obviously we have to consider that in cancerous ones higher viability percentages will be expected by using common viability assays, so the right results interpretation is fundamental. In fact, generally the lowest toxic concentrations, corresponding to the highest viability values in the dose response curves, match with the highest number of proliferative centers which accumulate the red vital dye NR. In this case, soon after the incubation with the vital dye (NR or MTT), assay data generated by the microplate reader recorded as Optical Densities (OD), i.e. readings at 540 nm, are filled into a Microsoft Excel spreadsheet and normalized.
Than it is calculated the mean of the absorbance values acquired from triplicate wells and normalized to the mean value of untreated control. The % of cell viability is calculated by the formula 100 – [(Abs540 treated cells/Abs540 untreated cells) × 100].
The data are based on three independent experiments. Finally, the % viability data will be transferred to GraphPad Prism software package (GraphPad Software Inc. v.5.00, San Diego, CA, USA) to plot the dose–response curves. The four-parameter logistic model by GraphPad Prism v.6.02 software required the input of Log transformed data to originate graphs, so the concentrations have to be ordinated with a logarithmic scale.
However, to overcome some misinterpretation, Prism software performs post tests following two-way ANOVA using the Bonferroni method. Such post test evaluate the linear trend and works by calculating linear regression on group mean (with data in Optical Density and not in % of viability) vs. column number, specifically by comparing each raw mean with control raw mean. This means that the dose response curves and the post test for trend have to be performed separately by using the same experimental data, but expressed differently as % of cell vitality over control and Optical Density (OD) for mean comparison respectively.
Can anyone tell me how to use Statistica to determine the LC50 for toxicity data? I'm interest in both generating the sigmoidal curve and determining the LC50 value.
Please read my paper 'Rath S, Sahu MC, Dubey D, NK Debata, RN Padhy (2011). Which values should be used as the lethal concentration 50 (LC50) with bacteria? Inter-Disciplin. Sci - Comput Life Sci. 3:138-143. (Springer)., in ResearchGate. If you find difficulties contact me.
Hi Dr. Irshad, if insisting in Excel there are a few options for you. I've attached an Excel spreadsheet to illustrate this:
1) (sheet 1) You can convert your m% response to probit values "=NORMSINV()", then do a linear regression vs log-converted concentration or dose.
2) (sheets 2 and 3) You can do non-linear regression through Solver. Solver is an add-in, built into Excel, that does least-squares regression... there are many models that can describe the sigmoidal shaped non-linear function... most common is the logistic model
Rath S, Sahu MC, Dubey D, NK Debata, RN Padhy (2011). Which values should be used as the lethal concentration 50 (LC50) with bacteria? Inter-Disciplin. Sci.─ Comput. Life Sci. 3:138-143. (Springer).
If you need a specific LD50/LC50 calculator, it is freely available at the following site: This calculator also provides estimation of 95% fiducial confidence limits for LD/LC values and is downloadable
Can the LD50 value change based on the concentration? I repeated an experiment with 3 times and LD50 values of the same drug are 7.8, 7.81 and 7.69 so it can be possible or not.
Thankfully the ic50.tk website is now working again but feel free to also try ic50.org. If you can give the tk site a few clicks, it might help the google ranking which has sunk after 4 days downtime!!