<!DOCTYPE html>
<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=UTF-8">
  </head>
  <body>
    <p> </p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm"><font
        face="FreeSans, sans-serif"><span style="font-variant: normal"><font
            color="#000000"><span style="text-decoration: none"><font
                size="2" style="font-size: 11pt"><span
                  style="font-style: normal"><b><span
                      style="background: transparent"><u>Please feel
                        free to share this offer for a PHD position at
                        IGMM in Montpellier</u><br>
                      <br>
                      Project:</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">DNA
                    large language models for end to end genome
                    interpretation</span></span></font></span></font></span></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><u><span
            style="font-variant: normal"><font color="#000000"><span
                style="text-decoration: none"><font size="2"
                  style="font-size: 11pt"><span
                    style="font-style: normal"><b><span
                        style="background: transparent">Motivation</span></b></span></font></span></font></span></u></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">Interpreting
                    the genome means modeling the relationship between
                    genotype and
                    phenotype, which is the fundamental goal of biology.
                    Achieving this
                    would revolutionize genetics, medicine, and
                    agro-tech. In clinical
                    genetics, it could lead to personalized treatments
                    tailored to each
                    patient's genome, enabling precision medicine. </span></span></font></span></font></span></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><u><span
            style="font-variant: normal"><font color="#000000"><span
                style="text-decoration: none"><font size="2"
                  style="font-size: 11pt"><span
                    style="font-style: normal"><b><span
                        style="background: transparent">Objectives</span></b></span></font></span></font></span></u></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">This
                    project fuses quantitative genetics with
                    bioinformatics and
                    cutting-edge Artificial Intelligence, using the
                    latest Deep Learning
                    Large Language Models for DNA to advance our ability
                    to predict the
                    phenotypes deriving from the observed genotypes.
                    This project is
                    based on the previous work of Dr. Raimondi on Genome
                    Interpretation
                    (GI) for the prediction of clinically relevant
                    phenotypes in humans. </span></span></font></span></font></span></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">Dr.
              Raimondi previous works focused on encoding
              Whole-Exome/Whole-Genome
              Sequencing (WES/WGS) data into compact, machine-readable
              features,
              while reducing overfitting caused by high-dimensional
              genomic data.
              However, to limit complexity, we had to rely on coarse
              gene-level
              summaries, such as mutational burden per gene, which
              sacrificed
              fine-grained genetic detail.</span></font></font></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">The
              goal of this project is to move beyond these gene-centric
              encodings
              by building neural network architectures that operate
              directly at the
              nucleotide level. To achieve this, we propose integrating
              pre-trained
              DNA LLMs as unsupervised feature extractors within GI
              models. These
              models, trained using self-supervised learning on entire
              genomes,
              capture rich patterns of DNA dependencies and can produce
              information-dense latent representations.</span></font></font></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">DNA
              LLMs have shown strong performance in various functional
              genomics
              tasks, such as identifying regulatory elements and variant
              effects.
              This project will evaluate whether their latent
              representations can
              improve phenotype prediction. The new DNA LLM methods will
              be
              prototyped on A.thaliana, which is a well known model
              organism. Later
              developments will be translated to the disease risk
              prediction of
              Inflammatory Bowel Disease (IBD). Unlike DNA LLM research,
              this work
              applies LLMs to the interpretation of individual-level
              WES/WGS data
              for disease risk, marking a novel use of these models in
              human
              genetic prediction.</span></font></font></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><u><span
            style="font-variant: normal"><font color="#000000"><span
                style="text-decoration: none"><font size="2"
                  style="font-size: 11pt"><span
                    style="font-style: normal"><b><span
                        style="background: transparent">Candidate
                        profile</span></b></span></font></span></font></span></u></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">We
                    are looking for a </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">motivated</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">and
                  </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">curious</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">candidate,
                    with a strong passion for science and for scientific
                    discovery
                    through the use and creation of new data science and
                    Machine Learning
                    methods. Bioinformatics and Genome Interpretation
                    are
                    multi-disciplinary and rapidly evolving fields.
                    Therefore, the
                    candidate is expected to 1) be eager to continuously
                  </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">learn</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">new
                    skills, methods and concepts, and 2) to </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">enjoy</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">finding
                    new solutions in the face of new and unforeseen
                    difficulties.</span></span></font></span></font></span></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">The
                    ideal candidate has </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">very
                      good</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">1)
                    python programming skills, 2) understanding of </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">the
                      mathematical foundations and principles</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">of
                    Machine Learning, Linear Algebra (vectorial and
                    matricial operations,
                    optimization), with a particular focus on </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">Neural
                      Networks</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">,
                    3) </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">problem
                      solving</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">skills,
                    4) familiarity with GNU/Linux environment and 5)
                    ability to
                  </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">multi-task</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">across
                    different projects. A good understanding of the
                    basic concepts of
                    Bioinformatics is not necessary but welcome. The
                    project will be
                    based on the development of </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">un-orthodox
                      Neural Network</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">models
                    with </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">Pytorch</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">.</span></span></font></span></font></span></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">B2
                    level of English is </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">required</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">.</span></span></font></span></font></span></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">The
              offer provides an initial 6-month contract, with
              possibility of
              renewal to 2 years. This project can be extended to 3
              years and
              offers the opportunity to obtain a PhD.</span></font></font></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><u><b><span
                        style="background: transparent">Research
                        environment</span></b></u></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><u><span
                  style="background: transparent">
                </span></u></span></font></span></font>
    </p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">The
                    recruited person will join the “AI for Genome
                    Interpretation”
                    team led by Dr. Daniele Raimondi at IGMM. The work
                    will be conducted
                    in an international (English-speaking) and
                    interdisciplinary
                    environment. </span></span></font></span></font></span></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">The
              Institute of Molecular Genetics of Montpellier (IGMM) is a
              joint
              research unit affiliated with the CNRS and the University
              of
              Montpellier. It comprises around 200 members, organized
              into 18
              research groups, 9 shared support services and 9
              technological and
              scientific platforms.</span></font></font></font></p>
    <p
style="font-variant: normal; font-style: normal; font-weight: normal; line-height: 100%; margin-bottom: 0.35cm; text-decoration: none">
      <font color="#000000"><font face="FreeSans, sans-serif"><font
            size="2" style="font-size: 11pt"><span
              style="background: transparent">IGMM
              is a multidisciplinary institute whose research has both
              fundamental
              and translational impact in molecular and cellular biology
              at the
              international level.</span></font></font></font></p>
    <p style="line-height: 100%"><a
        name="docs-internal-guid-b209c73f-7fff-de56-8ffe-0248c058f5d0"></a>
      <font face="FreeSans, sans-serif"><br>
        <span style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><u><b><span
                        style="background: transparent">Qualifications</span></b></u></span></font></span></font></span></font></p>
    <p
style="font-weight: normal; line-height: 100%; margin-bottom: 0.35cm">
      <font face="FreeSans, sans-serif"><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">The
                    ideal candidate has </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">very
                      good</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">1)
                    python programming skills, 2) understanding of </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">the
                      mathematical foundations and principles</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">of
                    Machine Learning, Linear Algebra (vectorial and
                    matricial operations,
                    optimization), with a particular focus on </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">Neural
                      Networks</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">,
                    3) </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">problem
                      solving</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">skills,
                    4) familiarity with GNU/Linux environment and 5)
                    ability to
                  </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">multi-task</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">across
                    different projects. A good understanding of the
                    basic concepts of
                    Bioinformatics is not necessary but welcome. The
                    project will be
                    based on the development of </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">un-orthodox
                      Neural Network</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><span
                style="background: transparent">
              </span></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">models
                    with </span></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><b><span
                      style="background: transparent">Pytorch</span></b></span></font></span></font></span><span
          style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="background: transparent">.
                    B2 level of English is required. Familiarity with
                    scientific
                    computing and libraries such as numpy, scikit-learn,
                    scipy, pytorch.</span></span></font></span></font></span></font></p>
    <p style="line-height: 100%"><font face="FreeSans, sans-serif"><br>
        <span style="font-variant: normal"><font color="#000000"><span
              style="text-decoration: none"><font size="2"
                style="font-size: 11pt"><span style="font-style: normal"><span
                    style="font-weight: normal"><span
                      style="background: transparent"><b>HOW TO
                        APPLY:</b>
<a class="moz-txt-link-freetext" href="https://umontpellier.nous-recrutons.fr/poste/7sqwj9n4vg-ingenieur-en-calcul-scientifique-ou-ingenieur-en-ingenierie-logicielle-fh/">https://umontpellier.nous-recrutons.fr/poste/7sqwj9n4vg-ingenieur-en-calcul-scientifique-ou-ingenieur-en-ingenierie-logicielle-fh/</a></span></span></span></font></span></font></span></font></p>
    <p style="line-height: 100%"><font face="FreeSans, sans-serif"><br>
      </font>-- <br>
    </p>
    <pre class="moz-signature" cols="72">Daniele Raimondi, PhD
Chaire de Professeur Junior CNRS
AI for Genome Interpretation group
Institut de Génétique Moléculaire de Montpellier (IGMM)
1919 Route de Mende
34090 Montpellier, France

-obscurum per obscurius-</pre>
  </body>
</html>