Tweet
Hello all,
I am a new student to bioinformatics and CLC genomics so would be grateful for any help given as I'm feeling slightly lost.
I have been given illumina data (fastq files) consisting of the genome sequences to a bacterial genome, initially to perform single- end RNA-Seq analysis and differential expression analysis.
My main focus is to find: The number of mutants.
The number of transposon insertions in total.
The number of transposon insertions per gene.
I have three conditions: Starting, first output and second output i.e. 3 groups.
I have replicate sequencing data for the first output and second output data however I only have one set of data for the starting.
Using CLC genomics, I have been able to trim my sequences, map them to an annotated genome and perform RNA-Seq analysis.
(An example to aid simplicity of my explanation. My experiment contains 1 starting library, 3 first output replicates and 3 second output replicates).
To start my comparison, I did a box plot of the three groups. This revealed that the individual samples all have a similar distributions, especially with respect to their own group although, the locations of the distributions differed. Because of this, I normalized my samples by quantile normalization which indeed made the samples comparable exemplified by a box plot showing each sample as having an equal distribution. Was this the right thing to do? Should I have chosen a different normalization method?
Next, I performed a statistical test on the proportions of the two groups using the starting library as the reference. I performed an unpaired Baggerley test of the three groups and chose the RPKM as the expression value to be used in the test.
(see manual here http://www.clcsupport.com/clcgenomic...oportions.html )
Now what I am unsure of the following things:
What is the easiest way to identify insertions into the genome through the data without individually checking the read mapped onto each gene?
What is the correct way to normalize my data?
Statistically what is the best test to use?
What values are most important for my test?
Should I use the fold-change, total gene reads or unique gene reads instead of the RPKM for the expression value to be used in the test?
Thanks in advance
I am a new student to bioinformatics and CLC genomics so would be grateful for any help given as I'm feeling slightly lost.
I have been given illumina data (fastq files) consisting of the genome sequences to a bacterial genome, initially to perform single- end RNA-Seq analysis and differential expression analysis.
My main focus is to find: The number of mutants.
The number of transposon insertions in total.
The number of transposon insertions per gene.
I have three conditions: Starting, first output and second output i.e. 3 groups.
I have replicate sequencing data for the first output and second output data however I only have one set of data for the starting.
Using CLC genomics, I have been able to trim my sequences, map them to an annotated genome and perform RNA-Seq analysis.
(An example to aid simplicity of my explanation. My experiment contains 1 starting library, 3 first output replicates and 3 second output replicates).
To start my comparison, I did a box plot of the three groups. This revealed that the individual samples all have a similar distributions, especially with respect to their own group although, the locations of the distributions differed. Because of this, I normalized my samples by quantile normalization which indeed made the samples comparable exemplified by a box plot showing each sample as having an equal distribution. Was this the right thing to do? Should I have chosen a different normalization method?
Next, I performed a statistical test on the proportions of the two groups using the starting library as the reference. I performed an unpaired Baggerley test of the three groups and chose the RPKM as the expression value to be used in the test.
(see manual here http://www.clcsupport.com/clcgenomic...oportions.html )
Now what I am unsure of the following things:
What is the easiest way to identify insertions into the genome through the data without individually checking the read mapped onto each gene?
What is the correct way to normalize my data?
Statistically what is the best test to use?
What values are most important for my test?
Should I use the fold-change, total gene reads or unique gene reads instead of the RPKM for the expression value to be used in the test?
Thanks in advance
Comment