Biology Reference
In-Depth Information
3.6 Annotation and
Functional Analysis
BLAST2GO is a Java interface tool with multiple features for
annotation and analysis of both nucleotide and protein sequences.
A detailed description on the capabilities and uses of the tool can
be found in the original publication [
31
]. Here, we focus on the
steps for sequence alignment and functional annotation.
1. Load the FASTA file containing the protein sequences.
2. Run alignment step by selecting “Blast” in the main control
panel. Then, select the desired database (nr or SwissProt),
number of Blast hits to be reported per sequence, minimum
Blast expected value, Blastp as the algorithm, and minimum
HSP length. We recommend using a minimum expect value
equal to 1e
−10
and minimum HSP cutoff equal to 33 and check
the “Low complexity filter” option.
3. After alignment step is done and all sequences received a
description, run mapping step to obtain gene ontology infor-
mation for all the sequences and start the annotation step.
4. To create a pie chart with the distribution of the mapped GO
terms for the dataset, select “Analysis” and “Make a combined
graph” option (
see
Notes 10
and
11
).
3.7 Clustering
Analysis
A common strategy applied to analyze large-scale data is to cluster
proteins accordingly to their expression profile. We use a freely
available program called PermutMatrix [
32
] to carry out the clus-
tering analysis.
1. In a Microsoft Excel worksheet, input raw spectral count val-
ues such that the columns represent samples and rows repre-
sent proteins.
2. Replace the missing values (i.e., spectral counts equal to zero)
by the mean of the biological replicates (
see
Note 12
).
3. Calculate the mean of the spectral count across the biological
replicates and normalize the data by dividing each biological
replicate mean by the mean of all values for a protein. This
ratio-based value should then be submitted to the base 2 loga-
rithmic transformation.
4. Prepare a standard tab-delimited text file containing the trans-
formed data matrix.
5. After loading the file in PermutMatrix, select Pearson distance
for the calculation of the dissimilarities. Then, select the aver-
age linkage (UPGMA) method for aggregation procedure.
6. After the result is generated, change the display options for a
better view of the data. If necessary, try a different clustering
method.
7. Export the tree by selecting “Export to Clipboard” and save
as a JPEG or Bitmap file.
