JEFFERIES GLOBAL HEALTHCARE CONFERENCE

CHARLES BAUM, M.D., PH.D., PRESIDENT AND CEO

JUNE 2015

A targeted oncology company developing a pipeline of cancer therapies for select patient populations.

Safe harbor statement

2

Certain statements contained in this presentation, other than statements of fact that are independently verifiable at the date hereof, contain “forward-looking” statements, within the meaning of the Private Securities Litigation Reform Act of 1995, that involve significant risks and uncertainties. For more detailed disclosures and discussions regarding such forward looking statements, please refer to Mirati’s filings with the U.S. Securities and Exchange Commission (“SEC”), including without limitation Mirati’s filings on Forms 10-K, 10-Q, and 8-K. Forward looking statements are based on the current expectations of management and upon what management believes to be reasonable assumptions based on information currently available to it. Such statements can usually be identified by the use of words such as "may," "would," "believe," "intend," "plan," "anticipate," "estimate," “expect,” and other similar terminology, or by statements that certain actions, events or results "may" or "would" be taken, occur or be achieved. Such statements include, but are not limited to, statements regarding Mirati’s development plans and timelines, potential regulatory actions, expected use of cash resources, the timing and results of clinical trials, and the potential benefits of and markets for Mirati’s product candidates. Forward looking statements involve significant risks and uncertainties and are neither a prediction nor a guarantee that future events or circumstances will occur. Such risks include, but are not limited to, potential delays in development timelines or negative clinical trial results, reliance on third parties for development efforts, changes in the competitive landscape, changes in the standard of care, as well as other risks described in Mirati’s filings with the SEC. We are including this cautionary note to make applicable, and to take advantage of, the safe harbor provisions of the Private Securities Litigation Reform Act of 1995 for forward-looking statements. The information in this presentation is given as of the date above and Mirati expressly disclaims any obligation to update or revise any forward-looking statements, whether as a result of new information, future events or otherwise, unless required by law.

Proven Targeted Oncology Strategy

Accelerated Development Approach

• Targeting genetic and epigenetic drivers of cancer

• Treating intrinsic and acquired mechanisms of resistance to targeted therapy

• Experienced and proven management team

• Creative and agile clinical development

ACCELERATED APPROVAL

FULL APPROVAL

Expanded indications

Genetically-defined patients

Mirati develops molecularly targeted oncology therapies for patients with cancer-driving genomic alterations

3

4 All programs owned by Mirati except certain Asian rights to Mocetinostat – Partnered with Taiho.

CANDIDATE/ INDICATION

PRIMARY TARGETS PRECLINICAL PHASE I PHASE 1b -

PHASE 2 REGISTRATION

TRIAL

MGCD265 Non Small Cell Lung Cancer (NSCLC) MET

Axl MGCD265 Solid tumors

MGCD516 Solid tumors

RET DDR Trk

Mocetinostat Bladder Cancer

HDAC 1,2,3,11 Mocetinostat

Diffuse Large B-Cell Lymphoma (DLBCL)

KIN

ASE

PRO

GR

AMS

EPIG

ENET

IC

PRO

GR

AM

Mirati is advancing multiple targeted oncology clinical pipeline programs

MGCD265 Multi-Targeted Kinase Inhibitor

6

MGCD265

MET plays a key role in NSCLC

*Raghav KPH, et al, Translational Lung Cancer Research 1.3 (2012):179-193. **Mirati data on file and Paik PH., et al. Cancer Discovery. Online May 13, 2015. Last accessed May 14, 2015 at http://cancerdiscovery.aacrjournals.org/content/early/2015/05/13/2159-8290.

• Higher MET expression rates correlate with advanced stages of tumor progression and poor clinical outcomes*

• Historical efforts to target MET generated inconsistent results due to incorrect development strategies as well as molecular challenges – MET protein overexpression by IHC versus MET genetic alterations – Driver alterations are more effectively targeted by small molecules

MET’s ROLE IN CANCER

• MET is a driver of tumor growth when it is genetically altered and activated by point mutations, exon 14 deletions and gene amplification in NSCLC patients** – MET mutations transform cell lines – Inhibition of MET mutations in xenograft models leads to tumor regression – MET mutations do not overlap with other drivers (e.g., ALK, ROS, EGFRm)

MET IN NSCLC

7

MGCD265

• 3 – 4% of NSCLC (including exon 14 deletion mutations)

MBIP

∆ Juxtamembrane

Kinase TM SEMA

• ~1% of NSCLC

MET mutations:

Kinase TM LZ LZ

MET

AX

L

Kinase TM SEMA

exon 14 / 15

• 2 – 3% of NSCLC MET gene amplifications:

Axl rearrangements:

MGCD265 is a selective inhibitor of the MET and Axl family receptor tyrosine kinases (RTKs)

Single agent MGCD265 targets genetic drivers in up to 8% of NSCLC patients

Kong Beltran et al Cancer Res 2006; Mitsudomi et al J Thoracic Oncol 2009; Seo et al Genome Res 2012.

8

MGCD265

ASCO 2015 abstract # 2589.

MGCD265 preclinical data in MET mutation and MET gene amplification xenograft models increase our confidence in targeted approach

MET Exon 14 Deletion PDX Lung Cancer Model

Start of treatment

MET Amplified Gastric Cancer Model

1000

2000

3000

0 8 16Treatment Day

Tum

or v

olum

e (m

m3 )

04 12

MGCD265 60 mg/kg QD

Vehicle

Start of treatment

Vehicle

MGCD265 60mg/kg QD

Start of treatment

Day Post Tumor Implantation

9

MGCD265

MGCD265 is differentiated by its unique binding mode and inhibition of Axl

MGCD265’s Unique Binding Mode:

• Allows for inhibitory activity against a broader range of MET alterations (including exon 14 deletions)

• Potential to avoid acquired resistance associated with secondary MET mutations

• Binds to kinase active site using induced fit and key intramolecular interactions within a deep hyrdophobic pocket

MGCD265

Other MET inhibitors

MET gene mutations (~3-4% NSCLC)

MET gene amplification (~2-3% NSCLC)

Axl Fusions (~1% NSCLC)

10

MGCD265

Phase 1 dose escalation study of MGCD265 achieved >90% inhibition of targets, and MTD of 1050 mg BID was well tolerated

Part 2: Dose expansion

• Patients with genetic alterations in MET or Axl

Cohort 1 N = 3 600mg BID

Cohort 2 N = 6 1200mg BID

Cohort 3 N = 3 1050mg BID

1050mg BID MTD / RP2D

• 2 DLTs • Grade 3 fatigue • Grade 3 diarrhea

• No DLTs

ClinTrials.gov identifier: NCT00697632.

11

MGCD265

MGCD265 dose expansion study is ongoing

Expansion Cohort 2 2nd-line Solid tumors (e.g., gastric cancer)

Expansion Cohort 1 2nd-line NSCLC

NGS screen for patients with:

•MET mutations •MET gene amplifications •Axl rearrangements

NSCLC

Other solid tumors

ASCO 2015 abstract #2589. ClinTrials.gov identifier: NCT00697632.

Initial clinical data confirm our targeted approach, resulting in tumor regression in NSCLC patients selected for MET genetic alterations.

12

MGCD265

MGCD265: 76-year old male, NSCLC with large lung tumor invading chest wall and ribs

Baseline: 01/15/2015

First Scan Post-Treatment: 02/23/2015

• MET exon 14 deletion mutation

• Extensive refractory NSCLC with metastases to the lung, retroperitoneal and retrocrural lymph nodes, malignant pleural effusion

• Received platinum-based chemotherapy followed by PD-L1 inhibitor therapy. – Best response to each treatment was progressive disease

• First scan after treatment with MGCD265: Cavitation of the lung/retroperitoneal mass, with resolution of pain and cough

ASCO 2015 abstract #2589.

MGCD265

MGCD265: 70-year old female, NSCLC with liver metastases

ASCO 2015 abstract #2589. 13

Baseline: 12/15/14

First scan post-treatment: 1/29/15

• MET exon 14 deletion mutation

• Liver metastases and right pulmonary lesion

• Underwent neoadjuvant chemoradiation followed by radical pneumonectomy for T3N1 disease, received chemoradiation for pleural and bone metastases, Stereotactic Body Radiation Therapy (SBRT) for pulmonary metastasis and resection of recurrence in the colon

• First scan after treatment with MGCD265: Extensive tumor necrosis and regression

14

MGCD265

MGCD265: 51-year old female, NSCLC with large lung tumor, bone and brain metastases

• MET amplification

• Refractory metastatic lung cancer with extensive lung disease

• Received EGFR inhibitor, chemotherapy and whole brain radiation

• First scan after treatment with MGCD265: Tumors in the lung showed regression

Baseline: 3/5/2015

First scan post-treatment: 4/13/2015

ASCO 2015 abstract #2589.

Initial data show preliminary evidence of clinical efficacy in heavily pretreated NSCLC patients with MET gene alterations

15

CONCLUSIONS

• MGCD265 is tolerable at the MTD of 1050 mg BID. Exposures achieved with this dose result in >90% inhibition of MET and Axl based on preclinical predictions and the biomarkers sMET and sAxl

• The first 3 patients (two with MET exon 14 deletion mutations and one with MET amplification) demonstrated tumor regressions

• Enrollment continues: – Patients with NSCLC with MET exon 14 deletions or MET gene

amplification – Patients with other solid tumors with genetic alterations of interest

16

MGCD265

We anticipate initiating the Phase 2 registration-enabling study for MGCD265 monotherapy in NSCLC by the end of 2015

Tissue or blood NGS screen for patients with:

• MET mutations • MET gene amplifications

• Open label, parallel-arm study of MGCD265

• Target Population: Recurrent or metastatic NSCLC with activating genetic alterations of MET with at least one prior treatment with a platinum based combination therapy

• Multinational: U.S., Canada, Europe and Asia

• Primary Endpoint: Objective Response Rate (ORR)

• Secondary Endpoint: Progression Free Survival (PFS)

We believe ORR >40% could lead to accelerated approval in 2nd-line setting for NSCLC followed by a confirmatory trial in 1st-line.

2nd-line NSCLC: MET mutations

2nd-line NSCLC: MET gene amplifications

17

MGCD265

MGCD265 has significant single agent NSCLC market opportunity

Mirati data on file; TCGA; AACR 2015 abstract #1118.

NSCLC Driver Alterations

MET amplifications

MET mutations

Axl rearrangements 1%

3.5%

3%

ALK4%

BRAF4%

ROS1%Her21%

MET amp 3%

MET mut 3.5%

Axl 1%

Other & Unknown42%

KRAS25%

EGFR15%

US NSCLC Annual Incidence: 191,000 Target population: 15,300

MET and Axl genetic alterations comprise up to 7-8% of NSCLC

• MET mutations: ~3.5% • MET amplifications: ~3% • Axl rearrangements: 1% Axl rearrangements 1%

T790M+ 7.5%

T790M - 7.5%

18

MGCD265

• Resistance is mediated through mutation and/or overexpression of alternative RTK targets and pathways, including MET and Axl

• Research shows that EGFR kinase inhibitor resistance can be reversed in vivo by combined EGFR and MET/Axl inhibition, a finding that validates combination therapy with EGFR and MET/Axl inhibitors to address therapeutic resistance

All tumors eventually become resistant to EGFR therapy

MGCD265 plus an EGFR/T790m inhibitor combination could target the majority of EGFR resistance in NSCLC

*Turke AB, et al. Cancer cell 17.1 (2010): 77-88., Zhang Z. et al. Nature Genetics 44.8 (2012): 852-862., Byers L.A., Diao L., Wang J., et al. Clinical Cancer Res 19. (2013): 279-290. Published online first Oct. 22, 2012.

Adapted from Gibbons, DL. Cancer Discov, 2014

19

MGCD265

Single agent MGCD265, and in combination with a 3rd-generation EGFR inhibitor, could treat a significant NSCLC patient population

NCI SEER, Mirati Data on File. *30% in Asia.

Annual NSCLC Incidence

191,000**

AXL 1%

MET 7%

EGFRm >15%*

EGFRm positive

MET and/or Axl positive

T790m positive

T790m negative

• MGCD265

• Initial focus: Accelerated approval for 2nd-line, followed by 1st-line

• Patients:15,300

• T790m inhibitor

• Initial focus: Accelerated approval for 2nd-line

• Patients: ~14,300

• MGCD265 + T790m inhibitor

• Initial focus: Accelerated approval for 2nd-line

• Patients: ~14,300

• T790m inhibitor

• MGCD265 + T790m inhibitor

• 1st-line expansion potential

• Patients: 28,650

15%

8%

50%

50%

20

MGCD265

Single agent MGCD265 has the potential for accelerated approval in NSCLC, gastric cancer and in combination with EGFR inhibitor

2015 2016 2017 2018 Si

ngle

Age

nt

ACCELERATED APPROVAL

2nd Line NSCLC Dose Expansion POC

in NSCLC & Solid Tumors

NSCLC: 2nd Line Phase 2 Registration Trial (MET mutations and MET gene amplification)

Combination Dose Expansion POC in NSCLC

Combination Study NSCLC 2nd Line Trial C

ombi

natio

n w

ith T

790m

in

hibi

tors

NSCLC: Confirmatory Randomized Trial in 1st Line for Full Approval

Solid Tumors (e.g. gastric): 2nd Line Phase 2 Registration Trial

NDA SUBMISSION

2nd Line NSCLC

MGCD516 Multi-Targeted Kinase Inhibitor

22

MGCD516

POTENTIAL INDICATIONS

TRK

Kinase JM Ligand Binding

Kinase JM TPM3

Kinase JM TPR

Kinase ETV6 • Mutations in NSCLC (~1%)

TrkB/C mutations

TrkA/C fusions

RET

D

DR

• NSCLC fusions (2%)

• NSCLC (1%) (including 3-4% lung SCC)

Kinase Coiled Motor

Kinase Motor

Kinase JM Discoidin

KIF5B RET fusions

DDR2 mutations

Single agent MGCD516 targets genetic alterations in up to 4% of NSCLC patients

An et al 2012, Ding et al 2008, Doebele et al, 2014; Eguchi et al 1999, Euhus el al 2002, Greco et al 2010, Hammerman et al 2011, Harada et al 2011, Kohno et al 2012, Lipson et al 2012, Marchetti et al 2008, Phay et al 2010, Sheng et al 2001

23

MGCD516

Dose escalation study with MGCD516 in advanced solid tumors is ongoing

ClinTrials.gov identifier: NCT02219711.

Part 2: Dose expansion • NSCLC patients with genetic alterations in RET, DDR or Trk identified by NGS

• Expected to begin in second half of 2015

Cohort 1 10 mg QD

Cohort 2 20 mg QD

Cohort 3 40 mg QD

Cohort 4 80 mg QD

Cohort 5 110 mg QD

Unselected patients

Upon MTD

Mocetinostat Spectrum Selective Histone Deacetylase (HDAC) Inhibitor

25

MOCETINOSTAT

Indication Regimen Phase Patient Selection Strategy

DLBCL Single-agent

Phase 2 Ongoing

• Refractory patients with genetic alterations in CREBBP and EP300

Bladder Single-agent

Phase 2 Ongoing

• Refractory patients with genetic alterations in CREBBP and EP300

Mocetinostat’s novel HDAC development strategy centers on epigenetic patient selection

There is growing evidence that HDACs may be able to increase the efficacy of immuno-oncology therapies.* We are exploring combinations with checkpoint inhibitors. We are also exploring combinations with other epigenetic agents as a potential next step in the mocetinostat development program.

Sing

le A

gent

Po

tent

ial

Com

bina

tion

*Maeda T, et al. Blood, vol. 96., Skov et al. Cancer Research, vol.65,2005., West AC et al. Cancer Res 2013.

26

MOCETINOSTAT

Gui Y et al Nature Genetics 2011.

CH1 KIX Bromo CH2 CH3 CREBBP Acetyltransferase

EP300 Bromo KIX CH1

Splice site Insertion or deletion Nonsense Missense

CH2 CH3

CREBBP and EP300 are inactivated through loss of function mutations in a unique spectrum of cancers

Mocetinostat demonstrated complete tumor growth inhibition in non-clinical models exhibiting CREBBP and EP300 mutations

• Histone acetyltransferases (HATs), including CREBBP and EP300, are associated with gene transcription and are counterbalanced by HDACs, which are associated with gene silencing.

• Through their complimentary roles, they regulate the expression of other genes involved in cell growth, survival and differentiation.

27

MOCETINOSTAT

ClinicalTrials.gov Identifier: NCT02236195.

IF > 3 RESPONSES STAGE 1 N=15 *Patients

STAGE 2 N=33 *Patients

STAGE 3 N=100 *Patients

IF > 9 RESPONSES

Mocetinostat Phase 2 study in selected bladder cancer patients is ongoing

*Patients have inactivating mutations of histone acetyltransferase genes CREBBP and EP300.

28

MOCETINOSTAT

Sponsored Phase 2/3 Registration Study

N = 150 Patients

N = 27 *Patients

Cohort 2: Follicular Lymphoma

Cohort 1: DLBCL (Relapsed/refractory)

N = 27 *Patients

ClinicalTrials.gov Identifier: NCT02282358.

Investigator-sponsored mocetinostat Phase 2 study in selected Non-Hodgkins lymphoma patients is ongoing

*Patients have inactivating mutations of histone acetyltransferase genes CREBBP and EP300.

Summary

30

2015 is poised to be a transformational year for Mirati

*As of March 31, 2015.

PROGRAM UPDATE ESTIMATED TIMEFRAME

MGCD265

Provide additional data on efficacy and duration of response from Phase 1/1b study ONGOING

Initiate Phase 2 registration-enabling study END OF 2015

MGCD516 Identify optimal dose MID - 2015

Initiate dose expansion cohorts for Phase 1/1b study 2ND HALF - 2015

Mocetinostat Data from ongoing Phase 2 trials ONGOING

FIN

ANC

IAL*

• NASDAQ MRTX

• Shares Outstanding*: 16.1M

• Cash*: $68M