RNA molecule. This step removes read 2s and bins read 1s into
separate files for the plus and minus strands of the genome (see
Fig. 2). On the command line, use the perl script Split_strands_
R1_only.plx to create a bash script for each library that will
remove read 2 and divide each bam file into separate plus and
minus strand files.
2. On the command line, use the bash command to execute each
bash file. This will produce two files for each library: one
containing each read 1 mapped to the plus strand of the
genome and the other containing each read 1 mapped to the
minus strand of the genome.
3.10 Determine Read
Depth for each 5
0 End
1. 1. You will now create text files reporting the number of times
that the first nt of read 1, representing an RNA 5
0 end, mapped
to each coordinate on each strand of the genome (see Fig. 2).
On the command line, for each read 1 bam file, use the bedtools genomecov command with the Àd and À5 options to
obtain the first nt read depth for each coordinate. Save the
output as a text file. This command will require an additional
file specifying the size of the chromosome, which is simply a
text file containing the genome name and chromosome length,
for example, gi|118467340|ref|NC_008596.1| 6988209
2. If your experiment includes biological replicates, you may now
wish to combine coverage from each replicate into a single file
for each strand in which each replicate for each condition is
present as a column. For example, use the script Make_flat_file_
retain_0.py to make a combined file from three replicate converted (“phos”) and nonconverted (“mock”) libraries:
python Make_flat_file_retain_0.py WT_mock_rep1_sb_plus.coverage WT_mock_rep2_sb_plus.coverage WT_mock_rep3_sb_plus.coverage WT_phos_rep1_sb_plus.coverage WT_phos_rep2_sb_plus.
coverage WT_phos_rep3_sb_plus.coverage WT_plus_combined_coverage.txt
3. Use a text editor to add headers to each column of the text file
produced in step 2.
4. Normalize converted and nonconverted libraries to correct for
variation in total reads between libraries. We will arbitrarily
select one converted library replicate as a calibrator and normalize the other converted library replicates to it, and likewise
arbitrarily select one nonconverted library replicate as a calibrator and normalize the other nonconverted library replicates to
that. To calculate a normalization factor for library x, divide the
total read count of the calibrator by the total read count of
library x. Then, use the script Multiply_by_values_replace_
zeros.py to (1) replace all zero coverage values with 1 and
526
M. Carla Martini et al.
separate files for the plus and minus strands of the genome (see
Fig. 2). On the command line, use the perl script Split_strands_
R1_only.plx to create a bash script for each library that will
remove read 2 and divide each bam file into separate plus and
minus strand files.
2. On the command line, use the bash command to execute each
bash file. This will produce two files for each library: one
containing each read 1 mapped to the plus strand of the
genome and the other containing each read 1 mapped to the
minus strand of the genome.
3.10 Determine Read
Depth for each 5
0 End
1. 1. You will now create text files reporting the number of times
that the first nt of read 1, representing an RNA 5
0 end, mapped
to each coordinate on each strand of the genome (see Fig. 2).
On the command line, for each read 1 bam file, use the bedtools genomecov command with the Àd and À5 options to
obtain the first nt read depth for each coordinate. Save the
output as a text file. This command will require an additional
file specifying the size of the chromosome, which is simply a
text file containing the genome name and chromosome length,
for example, gi|118467340|ref|NC_008596.1| 6988209
2. If your experiment includes biological replicates, you may now
wish to combine coverage from each replicate into a single file
for each strand in which each replicate for each condition is
present as a column. For example, use the script Make_flat_file_
retain_0.py to make a combined file from three replicate converted (“phos”) and nonconverted (“mock”) libraries:
python Make_flat_file_retain_0.py WT_mock_rep1_sb_plus.coverage WT_mock_rep2_sb_plus.coverage WT_mock_rep3_sb_plus.coverage WT_phos_rep1_sb_plus.coverage WT_phos_rep2_sb_plus.
coverage WT_phos_rep3_sb_plus.coverage WT_plus_combined_coverage.txt
3. Use a text editor to add headers to each column of the text file
produced in step 2.
4. Normalize converted and nonconverted libraries to correct for
variation in total reads between libraries. We will arbitrarily
select one converted library replicate as a calibrator and normalize the other converted library replicates to it, and likewise
arbitrarily select one nonconverted library replicate as a calibrator and normalize the other nonconverted library replicates to
that. To calculate a normalization factor for library x, divide the
total read count of the calibrator by the total read count of
library x. Then, use the script Multiply_by_values_replace_
zeros.py to (1) replace all zero coverage values with 1 and
526
M. Carla Martini et al.