Jokovich
Posts: 1
Joined: Mon Feb 08, 2021 5:05 am
Occupation: Parent

Re: Am I doing my microbiology meta-analysis correctly?

Postby Jokovich » Mon Feb 08, 2021 5:06 am

Hello everyone,
I have been looking for this information for a long time, I was very surprised when I found it here.

MadelineB
Moderator
Posts: 830
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist

Re: Am I doing my microbiology meta-analysis correctly?

Postby MadelineB » Mon Feb 08, 2021 8:27 am

Hello Wisteria,

It's good that you asked your local science fair organizers about the literature review. It's also good that you plan to enter data in a spreadsheet. If you have time, please share a pdf of the spreadsheet when you've entered data for the first 2-3 studies.

As for the forest plot, that plot is intended to compare results from studies that basically conducted the same experiment (or clinical trial). So, for the same reasons you conclude it wouldn't really make sense to combine your data into a meta-analysis, it also would not make sense to summarize your data with a forest plot.

Not to despair ... from what you have described, and, depending on how you set up your spreadsheet, I expect that data will be useful to create some graphs to illustrate the effect of PNAs on genes (or plasmids). Again, I think I can help you choose how to graph the data in your spreadsheet.

However, you would need to be careful how you describe what you graph. Researchers usually do not make statements like "prove the fact ... ." This might be a subtle distinction, but researchers usually are careful to say something like "these results suggest that it is possible to use PNAs to silence genes on plasmids".

If you have already read the Science Buddies project guide, here's the link. The section on hypothesis is helpful for the terminology! https://www.sciencebuddies.org/science- ... ience-fair

I look forward to seeing a pdf of your spreadsheet when you've entered the data for the first 2-3 studies!

Wisteria
Posts: 11
Joined: Sun Jan 31, 2021 11:01 pm
Occupation: Student

Re: Am I doing my microbiology meta-analysis correctly?

Postby Wisteria » Mon Feb 08, 2021 5:43 pm

Hello Madeline,
I have extracted the information for the two studies I sent you and attached it to this message. Please let me know if there are any issues accessing the file! I included the study author, study year, concentrations tested, measurements that they reported in, name of the plasmid that they used, gene targeted, # expressing gene after treatment, and # expressing gene after treatment with the control group. It would be great if there was a way to express the data on a graph!

As for the terminology, thank you for the reminder! I realized that I do say "prove" quite a bit, so I'll be sure not to make that mistakes in my final presentation!

Thank you so much!
Attachments
Graph.pdf
(35.14 KiB) Downloaded 16 times

MadelineB
Moderator
Posts: 830
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist

Re: Am I doing my microbiology meta-analysis correctly?

Postby MadelineB » Wed Feb 10, 2021 11:10 pm

Hello Wisteria,

Thank you for your patience and for sharing the pdf of your spreadsheet. After reviewing the info in your spreadsheet and the two references, I suggest that you might want to reorganize the columns. You might actually use separate spreadsheets: one spreadsheet to record experimental results that are using CFU/ml as a measure of the effect of PNAs on the bacterial growth as in the 2006 study of the effects of PNAs on E. coli. The second spreadsheet could then have columns for recording the response for studies that report their results in in relative units. The results for these studies will need to be summarized separately so having separate spreadsheets seems to make sense.

I have uploaded a pdf of a draft spreadsheet to illustrate my suggestions for recording experimental results that are using CFU/ml as a measure of the effect of PNAs. As you can see, I've added columns to record info about the report, the bacterial species used and the (target) gene. In order to facilitate reading these data for statistical and/or graphical analyses, there is one row for the response at each dose concentration for each agent/drug/PNA tested in the experiments. Note that I'm not suggesting that results for one bacterial species will be combined with results for another. Data for each species can be separated for the analyses. It seems easier to keep track of your data entry to use the same spreadsheet for the same type of experiments.

Let me know if this approach makes sense. You can follow the same approach to create a spreadsheet for studies that report their results in in relative units. I'll be happy to review a pdf of that spreadsheet too. Let me know if you have more questions.
Attachments
Draft gene silencing data.pdf
(167.2 KiB) Downloaded 13 times

Wisteria
Posts: 11
Joined: Sun Jan 31, 2021 11:01 pm
Occupation: Student

Re: Am I doing my microbiology meta-analysis correctly?

Postby Wisteria » Sat Feb 13, 2021 5:01 pm

Hello Madeline,
Thank you for the suggestion! It makes sense to separate the two reporting methods, but would it be possible to combine them if I converted the raw data into percentages, or is that not advised? Also, is there any way to report this on a visual graph rather than just a spreadsheet? This approach does sound like a great method though, so I'll be working on it!

Thank you so much for your help!

MadelineB
Moderator
Posts: 830
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist

Re: Am I doing my microbiology meta-analysis correctly?

Postby MadelineB » Sat Feb 13, 2021 6:24 pm

Hi Wisteria,

I suggest that you record the raw data in the spreadsheets. Once you have the raw data in the spreadsheets, you can see if it makes sense to convert the data to percentages or whatever. The spreadsheets will be the input "data" for the graphs.

Also, keep an open mind while you are recording the data in the spreadsheets ... maybe some of the studies don't report their results in a way that fits into either of the two spreadsheets. In that case, design a third spreadsheet to fit the third type of studies!

Be sure to keep a separate log to record the publication doi and assign each study an unique ID! That will be important to keep track of when you start to graph the data.

I'll be glad to look at the pdf of any spreadsheet you design for each type of study. And feel free to ask more questions!

Madeline

Wisteria
Posts: 11
Joined: Sun Jan 31, 2021 11:01 pm
Occupation: Student

Re: Am I doing my microbiology meta-analysis correctly?

Postby Wisteria » Mon Feb 15, 2021 1:46 pm

Hello Madeline,
Great, thank you so much! One last question - since this is the input data for the graph, how will I create the graph? Do I have to draw it or is there a software that can take my data and turn it into a graph?

Thank you!

MadelineB
Moderator
Posts: 830
Joined: Fri Jun 20, 2014 4:42 pm
Occupation: Biostatistician/Data Scientist

Re: Am I doing my microbiology meta-analysis correctly?

Postby MadelineB » Mon Feb 15, 2021 11:12 pm

Hello Wisteria,
Most if not all computer applications that produce graphs will read data from spreadsheets. But the important thing for your project is that your spreadsheets are set up to record the data from the studies you review so you can select which studies are included in which graph. From my limited review of the two studies, it would not make sense to try to create a single graph to show the results of those two studies. That is why I suggested creating at least two spreadsheets to accommodate the different types of experiments represented by those studies. Think of the spreadsheets as lab notebooks for recording the type of experiment, design and results. If you record that information, then you can use the computer graphics application to prepare graphs for each type of experiment.

The number of spreadsheets and the data recorded in each will depend on the type of studies you see as you do your review. Hopefully you will find that the studies you are reviewing used similar experimental techniques and reported the results in similar formats. It will be helpful for you to learn if many studies used distinct experimental techniques. That could make it harder! I'll will be glad to review the spreadsheet designs!

Let me know if you have more questions!

Madeline


BTW, Tableau Public or Tableau Desktop are examples of computer graphics applications. Tableau Public is free, Tableau Desktop offers free one-year licenses to students who have an "edu" email.


Return to “Grades 9-12: Life, Earth, and Social Sciences”