Undergrad Research Experience

Spring 2021

Participants

• Eamen Ho: BIOL48002
• Afsana Rahman: BIOL48002
• Roman Shimonov: BIOL48002
• Zaheen Hossain: BIOL48002
• Ariel Cebelinski: BIOL48002
• Jean Ady: BIOL48001 & BIOl48002
• Mohamed Elgallad Volunteer
• Anh Pham: Volunteer
• Jannatul Ashpia: Volunteer

Reading List

• Design of broadly active anti-microbial vaccines:
  • Burton & Topol (2021). A commentary
  • Spensley et al (2019). Reverse Immunology. Scientific Report
  • Cohen et al (2021). Cocktail approach
  • an ASM review
• Group B Streptococcus (GBS) CC17 Evolution: Almeida et al (2017)
• CoV mutation analysis: Alouane et al (2020)
• HIV compartmentalized eolution: Evering et al (2014)

Project 1. Covid mutation analysis

• NCBI Covid genome files
• R Markdown file (by Janatul)

# Shared Afsana 
# Parsing NCBI SAR-CoV-2 genome and get gene length
#Parse through Covid Mutation File to obtain genes and their gene length. Columns: gene/matpeptide name, gene length, starting coordinate, ending coordinate
from Bio import SeqIO
import sys
inputFile=sys.argv[1] #multiple files can work with code
genome_record = SeqIO.read(inputFile, "genbank") #read file for analysis
seenProduct={} #create dictionary
for feature in genome_record.features:
    if feature.type == "gene":   #specifically look for gene feature, then loop through it to find location and name of each gene.remember that "gene" is a type and its name is a qualifier.
        name= feature.qualifiers["gene"]
        loc  = feature.location
        coordinates=str(feature.location)
        length=loc.end - loc.start + 1
        loc_start= str(loc.start) #starting coordinate of gene
        loc_end=str(loc.end) #ending coordinate
        print(name[0] + '\t' + str(length) + '\t' + loc_start + '\t'+ loc_end)
    if feature.type == "mat_peptide": #  specifically looking for matpetide to obtain genes of proteins. set loop to find location and name of each gene. remember that "gene" is a type and its name is a qualifier.
        name= feature.qualifiers["product"]
        if name[0] in seenProduct:
            continue #skips first duplicate of gene.
        else:
            seenProduct[name[0]]=1
            gene=feature.qualifiers["gene"]
            loc  = feature.location
            #length=(loc.end - loc.start + 1)
            length = len(loc)  
            loc_start= str(loc.start)
            loc_end=str(loc.end)
            print(name[0] + "\t" + str(length)  + '\t' + loc_start + '\t'+ loc_end)

# Shared by Afsana
# UNIX COMMANDS W COVID-19 GENOME MUTATION SET (CREATE COUNT OF SYNONYMOUS & NON-SYNONYMOUS MUTATIONS FOR EACH GENE)
grep synonymous cov-snps.tsv | cut -f3 | sort | uniq -c
grep synonymous cov-snps.tsv | cut -f3 | sort | uniq -c > count_synoymous_covid
grep missense  cov-snps.tsv | cut -f3 | sort | uniq -c > count_missense_covid
paste count_missense_covid count_synoymous_covid 
paste count_missense_covid count_synoymous_covid | tr -s ' ' 
paste -d ' '  count_missense_covid count_synoymous_covid | tr -s ' '
paste -d ' '  count_missense_covid count_synoymous_covid | tr -s ' ' | sed ' '
paste -d ' '  count_missense_ccovid count_synoymous_covid | tr -s ' ' | sed "s/ //"
paste -d ' '  count_missense_covid count_synoymous_covid | tr -s ' ' | sed "s/ //" | cut -f 1-3 -d ' '
paste -d ' '  count_missense_covid count_synoymous_covid | tr -s ' ' | sed "s/ //" | cut -f 1-3 -d ' ' > count_mutations_covid.tsv
paste -d ' '  count_missense_covid count_synoymous_covid | tr -s ' ' | sed "s/ //" | cut -f 1-3 -d ' '

# Shared by Roman
# compare syn vs nonsyn
library(tidyverse)
library(ggrepel)
ggplot(data=covid, aes(x=Missense,y=Synonymous, label=Gene)) + geom_point() + geom_text_repel() + geom_smooth(method="lm")

# Shared by Eamen PYTHON CODE FOR COVID MUTATION COUNTS
import re
import sys
gene_name = {}
with open(sys.argv[1]) as file:
    for line in file:
        lines = line.split()
        if re.match("noncoding", lines[7]):
            continue
        if lines[2] in gene_name:
            mutation = lines[7].split(',')

            for item in mutation:
                if item == 'missense':
                    gene_name[lines[2]]['total'] += 1
                    gene_name[lines[2]]['missense'] += 1
                elif item == 'synonymous':
                    gene_name[lines[2]]['total'] += 1
                    gene_name[lines[2]]['synonymous'] += 1
                else:
                    continue
        else:
            gene_name[lines[2]] = {}
            gene_name[lines[2]]['total'] = 0
            gene_name[lines[2]]['missense'] = 0
            gene_name[lines[2]]['synonymous'] = 0

            mutation = lines[7].split(',')
            for item in mutation:
                if item == 'missense':
                    gene_name[lines[2]]['total'] += 1
                    gene_name[lines[2]]['missense'] += 1
                elif item == 'synonymous':
                    gene_name[lines[2]]['total'] += 1
                    gene_name[lines[2]]['synonymous'] += 1
                else:
                    continue

for gene in gene_name:
    print(gene +"\t"+ str(gene_name[gene]['missense']) +"\t"+ str(gene_name[gene]['synonymous'])
          +"\t"+ str(gene_name[gene]['total']))

Project 2. Group B STreptoccocus genomics

• GenBank submission of assemblies

Fall 2020

Participants

• Eamen Ho: Volunteer research assistant
• Ramandeep Singh: BIOL 48002
• Desiree Pante: BIOL 48001
• Afsana Rahman: Volunteer research assistant
• Roman Shimonov: BIOL 48002
• Justin Hiraldo: BIOL 48002
• Zaheen Hossain: Volunteer research assistant
• Jerry Sebastian: Volunteer research assistant
• Ariel Cebelinski: Volunteer research assistant

Schedule

• Tuesdays at 12 noon - 2pm by Zoom
• Sept 1, 2020. Week 1. Meet & Greet; Intro to projects
• Sept 8, 2020. Week 2. Presentations (background, data, and methods), based on assigned readings

Project 1. Structure & evolution of multipartite genome of Lyme disease bacteria

• Participants: Desiree & Ramon (Summer 2020), Jerry
• Readings
  • Review: deCenzo & Finan (2017).
• Data set: lp54 & cp26 plasmids
• TO DO:
  • Week 1. 9/8/2020, 12 noon: 5-slides presentation on multipartite bacterial genome evolution (based on the paper above)
  • Week 2. 9/15, 12 noon: Use prorgram codonO to calculate codon bias (SCUO) for replicons (n=23) on Borrelia burgdorferi B31 genome
  • Week 3. 9/22, 12 noon: codonO paper presentation (Jerry)

Project 2. OspC Cross-reactivity analysis

• Participants: Justin, Roman
• Readings: Ivanova et al (2009)
• Tool: ImageJ
• Data set (to be sent)
• To Do
  • Week 1. 9/8/2020 12 noon: 5-slide presentation on background, material & methods, and data capture using ImageJ
  • Week 2. 9/15: Create Excel sheet to capture immunoblot intensities on C3H mice & P.lucus. Capture background for each serum. Getting ready to makes plots in R/Rstudio

Project 3. Clostridium transcriptome analysis

• Participants: Eaman, Zaheen
• Readings
  • SPARTA pipeline
  • Fletcher et al (2018).
  • Gu et al (2018)
• Data set: posted on "genometracker.org"
  • Wild type transcriptome at 12 hour, paired-end read files:
  • /home/azureuser/18134XR-29-01_S0_L001_R1_001.fastq.gz
  • /home/azureuser/18134XR-29-01_S0_L001_R2_001.fastq.gz
• To Do
  • Week 1. 9/8/2020 12 noon:
    • A short presentation on C. diff transcriptome (one of the 2 papers above)
    • Demo on read quality using FastQC and mapping reads to reference genomes with bowtie
  • Week 2. Use HT-Seq to quantify RNA abundance for C. diff genes.
    • HTSeq installed
    • Try this protocol first
  • Commands

According to: reference; Bowtie website

bioseq -i'genbank' R20291.gb > ref.fa # make FASTA file
bowtie2-build ref.fa index # build index
# -S: sam output (otherwise bam) 
bowtie2 -x index -S 18134XR.sam -1 ../18134XR-29-01_S0_L001_R1_001.fastq.gz -2 ../18134XR-29-01_S0_L001_R2_001.fastq.gz
# ref.gff3: need to run sed "s/Chromosome/FN545816/"
# need to use "-i"; default is "gene_id"
conda activate qiulab # change environment to access htseq
htseq-count -m union --stranded=yes 18134XR-29-01.sam ~/xingmin-cdiff/ref.gff3 -i=Parent > 18134XR-29-01.counts
samtools view -b 18134XR-29-01.sam -o 18134XR-29-01.bam # compress sam file into bam file

Project 4. Protein classification using natural language processing

• Participants: Afsana & Ariel
• Goal: Classify protein sequences
• Week 1. 9/8/2020 Readings:
  • Rives et al (2019)
  • Lan et al (2019)
• Week 2. Find/Explore ALBERT resources & Tutorials
• Code from Hansaim Lim
• Transformer: Pretrained models in natural language processing
• DNAbert paper

DNAbert: github code

• Including Albert
• Google albert library: github

Sample BioPython script:

#!/usr/bin/env python

import sys
import json
from Bio import SeqIO

alnFile = sys.argv[1] # read file as the first argument
seqList = [] # initialize a list
for record in SeqIO.parse(alnFile, "fasta"):
    seqList.append({"id": record.id,
                    "seq": str(record[0:3].seq) # use the str() function to convert object to string
                }) # get residue2 1-3

print(json.dumps(seqList)) # print to JSON format
exit