IPython line magics¶
In [1]:
%pwd
Out[1]:
In [2]:
%cd ~/BMS270/data/
In [3]:
%ls
Bash commands¶
In [4]:
!ls
In [5]:
!head -c 64 sample_table.csv | hexdump -C
Coverting among hexadecimal, decimal, and strings in python
In [6]:
0x52
Out[6]:
In [7]:
chr(0x52)
Out[7]:
In [8]:
"\x52"
Out[8]:
In [9]:
ord("R")
Out[9]:
In [10]:
"%x" % ord("R")
Out[10]:
Reading a text file in python¶
In [11]:
fname = "sample_table.csv"
In [12]:
fname
Out[12]:
In [13]:
type(fname)
Out[13]:
In [14]:
fp = open(fname)
In [15]:
fp
Out[15]:
In [16]:
fp.readline()
Out[16]:
In [17]:
fp.readline()
Out[17]:
In [18]:
fp.readline()
Out[18]:
In [19]:
fp.readline().rstrip()
Out[19]:
In [20]:
fp.readline().rstrip().split(",")
Out[20]:
In [21]:
from csv import reader
In [22]:
r = reader(fp)
In [23]:
next(r)
Out[23]:
In [24]:
row = next(r)
In [25]:
row
Out[25]:
Other newline conventions¶
- The old MacOS convention was \r
- The DOS/Windows convention is \r\n
In [26]:
!head -c 64 sample_table.oldmac | hexdump -C
In [27]:
fname = "sample_table.oldmac"
fp = open(fname)
fp.readline()
Out[27]:
In [28]:
fp
Out[28]:
In [29]:
fp = open(fname,newline="\n")
In [30]:
fp.readline()
Out[30]:
In [31]:
!head -c 64 sample_table.windows | hexdump -C
In [32]:
fname = "sample_table.windows"
fp = open(fname)
fp.readline()
Out[32]:
In [33]:
fp = open(fname,newline="\n")
fp.readline()
Out[33]:
Reading a binary file in python¶
In [34]:
!head -c 64 SRR7541452.6M.fastq.gz | hexdump -C
In [35]:
fname = "SRR7541452.6M.fastq.gz"
In [36]:
fp = open(fname)
In [38]:
fp.readline()
In [39]:
fp = open("SRR7541452.6M.fastq.gz","rb")
In [40]:
fp
Out[40]:
In [41]:
fp.read(10)
Out[41]:
- 1f 8b is the gzip magic
- 08 specifies DEFLATE compression
- 00 means no flags set
- c9 cc aa 5c is MTIME
- 00 means no extra flags set
- 03 means generated on Unix (Linux)
In [42]:
fp.seek(0)
s = fp.read(10)
In [43]:
s
Out[43]:
In [44]:
s[4:8]
Out[44]:
In [45]:
import time
t = 0
for i in reversed(s[4:8]):
t = t << 8
t += i
t
time.strftime("%Y-%m-%d %H:%M:%s", time.localtime(t))
Out[45]:
In [46]:
fp.close()
In [47]:
import gzip
In [48]:
fp = gzip.open(fname)
In [49]:
fp
Out[49]:
In [50]:
fp.readline()
Out[50]:
In [51]:
fp.readline()
Out[51]:
In [52]:
fp.readline()
Out[52]:
In [53]:
fp.readline()
Out[53]:
In [54]:
fp = gzip.open(fname,"rt")
In [55]:
fp
Out[55]:
In [56]:
fp.readline()
Out[56]:
In [57]:
for i in range(10):
print(fp.readline())
Parsing Illumina base qualities¶
Every 4th line of the FASTQ file gives the per-base quality scores (probability of a miscall) encoded as single characters.
$P = 10^{\frac{-Q}{10}}$
$Q = ord(C)-q_0$
$q_0 = 33$ 64 for older Illumina files
In [58]:
q0 = 33
C = "D"
Q = ord(C)-q0
P = 10**(-Q/10)
P
Out[58]:
In [59]:
C = "I"
Q = ord(C)-q0
P = 10**(-Q/10)
P
Out[59]:
In [60]:
fp.seek(0)
header = fp.readline().rstrip()
print(header)
seq = fp.readline().rstrip()
print(seq)
header2 = fp.readline().rstrip()
print(header2)
quals = fp.readline().rstrip()
print(quals)
In [61]:
Qs = []
for C in quals:
Q = ord(C)-q0
Qs.append(Q)
print(Qs)
In [62]:
len(seq)
Out[62]:
In [63]:
%matplotlib nbagg
import matplotlib.pyplot as plt
In [64]:
fig = plt.figure()
plt.plot(Qs)
Out[64]:
In [65]:
for i in range(20):
header = fp.readline().rstrip()
seq = fp.readline().rstrip()
header2 = fp.readline().rstrip()
quals = fp.readline().rstrip()
Qs = []
for C in quals:
Q = ord(C)-q0
Qs.append(Q)
plt.plot(Qs)