Internship : Bridging Ex Vivo Drug Response Data with Real-World Multi-Omics Cohorts to Inform Translational Strategy in AML

In the context of a translational program in acute myeloid leukemia (AML), we aim to connect mechanistic insights derived from a small-scale ex vivo dataset (~30 patients) with a larger, deeply characterized real-world (RWE) AML cohort. The ex vivo dataset includes compound efficacy data in both monotherapy and combination settings, alongside pro- and anti-apoptotic marker expression and bulk RNA sequencing. In contrast, the RWE cohort contains no pharmacological data but offers rich multi-omics (RNA-seq, DNA-seq, mutational data) and extensive clinical annotation (treatments, outcomes, subtypes).

embedding, or profile matching. By computing per-patient scores of “sensitivity likelihood,” we aim to stratify RWE patients according to their projected pharmacological behavior and interpret these subgroups in light of clinical and genomic characteristics.

This bridging approach is methodologically challenging due to the small size of the ex vivo dataset. The intern will therefore explore and implement regularized models (e.g., Lasso, ridge), conservative validation techniques (e.g., bootstrapping, LOOCV), and biologically interpretable scoring systems that allow reliable knowledge transfer across datasets. Ultimately, this work will inform biomarker-driven hypotheses and support early go/no-go decisions by bringing mechanistic evidence into real-world population contexts.

This internship offers a high-impact opportunity at the interface of data science, translational oncology, and decision support in drug development.

The goal of this internship is to develop a methodological framework to transfer predictive knowledge from the ex vivo system into the RWE setting. This includes identifying robust transcriptomic signatures of drug sensitivity in the ex vivo data and projecting these signatures into the RWE cohort using similarity-based approaches such as gene set scoring, transcriptomic.

Description du profil :

Profile :

• Master’s degree (M2) or final-year engineering student in bioinformatics, biostatistics, data science, or related field.
• Proficient in Python or R, with experience in data analysis and modeling (machine learning, regression, similarity scoring).
• Hands-on experience with transcriptomic or multi-omics data, ideally RNA-seq.
• Familiarity with small sample modeling challenges or knowledge transfer techniques.
• Strong scientific communication skills, demonstrated through reports, presentations, or collaborative projects.
• Soft skills: team spirit, communication, assertiveness, organization, adaptability, and problem-solving.

Avantages :

• Télétravail : 4 à 8 jours par mois selon l’activité et le poste
  • Horaires variables individualisés
  • Accès gratuit à la salle de sport sur site (cours collectifs + appareils de musculation)
  • Navettes gratuites Servier (7 lignes)
  • Certification Well Platinum (Bien être des collaborateurs et préservation de l’environnement)

• Informations pratiques
• Site : Institut de R&D Servier Paris-Saclay, Gif-sur-Yvette (91)
• Durée : 6 mois
• Date de début : Jan 2026
• Type de contrat : Stage
• Rémunération : 1100 € - 1400 €
• Accessibilité : Site accessible aux personnes en situation de handicap 

Comment postuler ?

Envoyez-nous votre candidature. Vous aurez l’opportunité de rencontrer notre équipe lors d’un processus en deux étapes : un entretien avec nos managers, suivi d’un entretien avec nos référentes RH.

Nous sommes engagés pour l’égalité des chances et le développement des talents dans toute leur diversité. Nous accordons autant de valeur à l’expérience qu’à l’envie de s’engager au quotidien pour être utile au progrès thérapeutique au bénéfice des patients. Si vous vous reconnaissez dans cette offre et ces quelques lignes, saisissez cette opportunité de nous rencontrer !

#MovedByYou