taeyon998/novellaB_bioinformatics

My work as an undergraduate research assistant under professor Hongyi Xin.

novellaB_bioinformatics

My work as an undergraduate research assistant under professor Hongyi Xin in UMJI_SJTU.

##Task

1. Simulate and plot a "Median UMIs per cell vs. Raw Reads per cell" graph.

   • Language: MATLAB
   • Objective: See how many raw reads are needed to see at least x reads per UMI
   • Criteria:
     • Generate random RNA data (eg. 1024 molecules per UMI, 30 UMIs in total)
     • Apply capture rate 0.2%, which is the standard capture rate used nowadays for sc-RNA sequencing (2020)
     • See how many molecules were captured for each UMI (eg. 3 reads for 1st UMI, 1 read for 2nd UMI...)
     • Plot 2D graph, average reads per UMI vs reads (total captured molecules)

2. Plot the Binomial Distribution for "Number of captured UMIs per cell for different capture rates & number of copies"

   • Language: Python
   • Objective: See how likely it is to get at least N reads per UMI, given a capture rate and number of copies per UMI.
     Biologists could use this model to adjust the (1) PCR rate or (2) the capture rate in the sequencing machine.
   • Criteria:
   • Plot a 3D graph, (x reads per UMI, capture rate r) vs. probability of capturing x molecules)
   • Receive user input: (1) number of UMI, (2) number of molecules per UMI, (3) different capture rates
   • 

2.1 Given certain PCR cycle, find the minimum capture rate (r) needed, and vice versa, in order to achieve below a desired error rate. Error rate is defined as the chance of completely missing the UMI during sequencing.

• Language: Python
• Objective: Missing a UMI during sequencing could lead to a misinterpretation of the cell. There are 2 variables that influence this chance: (1) PCR amplification cycles, (2) capture rate. Also, there would be a (3) thershold probability of missing the UMI, in which the biologist would want. A biologist could say okay to, for instance, a 1% chance of missing the UMI.
• Criteria:
  • Plot a 3D graph of independent variables (1) PCR amplification cycles and (2) capture rate, and a dependent variable P(x=0), which is the probability of capturing 0 molecules per UMI (completely missing the UMI).
  • Input capture rate & threshold probability -> minimum PCR cycles needed
  • Input PCR cycles & threshold probability -> minimum capture rate needed
•