MAZTER-seq, el codigo de m6A de levadura a mamíferos, �y la experiencia de mi doctorado en Israel.
Dr. Miguel Garcia
Postdoc en el Laboratorio Schwartz Depto. Genética Molecular
Instituto Weizmann de Ciencia
Marzo 2022
Presentación para: Universidad Técnica de Ambato
Outline
MAZTER-seq�and the m6A code
m6A is everywhere
N6-Methyladenosine
Adenosine
METTL3
IME4
m6A technology took a leap with antibody-based high-throughput profiling
IP-based methods have several limitations
RNA
Additional limitations of IP-based m6A detection
Although there have been improvements to alleviate previous limitations, no protocol existed providing complex, single-nucleotide resolution m6A mapping and quantification from limited starting mRNA material.
2019:
MAZTER-seq an m6A quantification technology
Graphical Outline
RNA
Dr. Sarit Edelheit Flohr
ACA
ACA
mazF
MAZTER-seq quantitatively captures methylation levels
MAZTER-seq’s cleavage efficiency estimations are highly reproducible.
MAZTER-seq is quantitative
ACA
ACA
ACA
ACA
RNA
ACA
88 nt
Quantifying m6A in yeast
WT
IME4 Δ/Δ
Input
IP
Input
IP
rep1
rep2
rep1
rep2
rep1
rep2
rep1
rep2
m6A
m6A
Experimental design
Pair-wise correlation (MAZTER-seq data)
MAZTER-seq quantitations at known m6A sites
Distribution of cleavage efficiencies (y-axis) at known m6A sites in RNA extracted from WT ime4Δ/Δ strains with versus without m6A-IP treatment
mazF biases are removed in �Δ Cleavage efficiency
mazF Cleavage Efficiency is biased by its own consensus sequence and by secondary structure access
Biases are removed when calculating the difference between WT and KO, here on Δ Cleavage efficiency
De novo m6A detection
WT
IME4 Δ/Δ
Input
IP
Input
IP
rep2
rep1
rep2
rep1
rep2
rep1
rep2
m6A
m6A
1
2
3
We used three comparisons to detect m6A sites with high confidence.
De novo m6A detection
Distribution of m6A-seq sites across the confidence groups defined via MAZTER-seq
Distribution of m6A-seq scores from (Schwartz et al., 2013) by MAZTER-seq confidence groups.
SCARLET validation of new m6A sites.
Validation rate 12 /14
M.Sc Ursula Toth
Prof. Walter Rossmanith
Medical University of Vienna
The m6A code
Model-derived coefficients value for each nucleotide at a given position
Relative importance of variables. Difference in the full model R2 when removing each of the variables one-in-one-out.
The m6A code
Variability of methylation levels can be predicted via local sequence and secondary structure information in both new and previously detected sites.
SCARLET quantifications and m6A levels predictions highly correlate
Model is able to predict m6A sites
Top predicted sites have a clear separation from background in m6A-seq derived data (Schwartz et al., 2013).
SCARLET validation of high-scoring predicted sites (5 out of 5)
m6A-Seq score
MAZTER-seq in mouse
1
MAZTER-seq in mouse
MAZTER-seq derived stoichiometries in Mouse Embryonic Stem Cells
Stoichiometries are highly predictable by sequence information
New sites in mouse
MAZTER-seq detection groups
Antibody-based measurements show enrichment in detection groups
m6a-seq IP dependent detection is influenced by its stoichiometry
m6A-seq antibody-based detection is influenced by stoichiometries as an stochastic sampling
The m6A code in mouse
Similarly to the yeast model. The mouse stoichiometry model favors A in -4, G at -2 and -1, and T in +3 and +4 positions
Relative importance of variables
The m6A code is conserved
Yeast
The code is conserved across species
Mouse
MAZTER-seq is able to quantify m6A across genetic perturbations
FTO over expression nor deletion seem to affect m6A levels
ALKBH5 over-expression reduces m6A levels��METTL3 deletion dramatically decrease m6A levels
Summary
Mi experiencia de doctorado en Israel
Aplicar a un doctorado internacional
Ventajas competitivas al aplicar
Proceso de adaptación
Vida académica �y logros
Vivencias
Algunas herramientas del científico computacional moderno
Inglés
Programación
Programación�Python y R
Técnica pomodoro y KanbanFlow
Toda � raba!
Contacto