Fasta to json via python dictionary

Question

[biology.may20]

Good day,

Could you please help with my code: I need to convert from fasta file to json-object and I'm doing this via python dictionary. I have stock with two things:

from Bio import SeqIO

from Bio.SeqIO import parse

my_dict = {}

with open("Salm_ser_Enteritidis.fasta", 'r') as new_fasta:

    for x in SeqIO.parse(new_fasta, 'fasta'):

        dataset = x.id

        sequence = x.seq

        my_dict = {"dataset": x.id,

                 "sequence": x.seq}

print(my_dict)

Output is

{'dataset': 'PYJS01003085.1', 'sequence': 
Seq('TCTCATCCGCCAAAACATCTTCGGCGTTGTAAGGTTAAGCCTCACGGTTCATTA...TTA')

but I need only actial sequence without word Seq for key 'sequence', so that output was:

{'dataset': 'PYJS01003085.1', 'sequence': 

'TCTCATCCGCCAAAACATCTTCGGCGTTGTAAGGTTAAGCCTCACGGTTCATTA...TTA'

and the second problem:

import json
my_json = json.dumps(my_dict)

output TypeError: Object of type Seq is not JSON serializable

but via print(type(my_dict)) I have output that my_dict is <class 'dict'>, not type Seq. How to convert to dict for further # json?

Thanks in advance, Galina

[LauferVA]

please see: https://stackoverflow.com/questions/7100125/storing-python-dictionaries

Answer 1

[Carambakaracho]

Hi, you generate a SeqRecord Object and the seqrecord.seq is a seq object, not a string. You can convert it to a string with an explicit string cast (conversion) via str(seqrecord.seq), see here

from Bio import SeqIO
import json

# from Bio.SeqIO import parse  # you won't need that, you imported the whole SeqIO package above

my_dict = {}

with open("Salm_ser_Enteritidis.fasta", 'r') as new_fasta:
    for x in SeqIO.parse(new_fasta, 'fasta'):
        # dataset = x.id
        # sequence = x.seq # you won't need that, if the variables go out of scope and aren't used
        my_dict = {
            "dataset": x.id,
            "sequence": str(x.seq) # string cast to SeqRecord seq object
        }

my_json = json.dumps(my_dict)  # maybe use json.dump directly to a file (use pretty print option, eg. indent=2)?

json.dump reference

[biology.may20]

thank you a lot :)

I've used dump and now it's looks awsome !))

with open('my_dict.json', 'w') as f: json.dump(my_dict, f, indent=2)

[LauferVA]

ya the other functions in the link i posted should be helpful afterwards as well ..